File size: 136,165 Bytes
17e971c 842d8e8 17e971c 842d8e8 17e971c 842d8e8 17e971c 842d8e8 17e971c 842d8e8 17e971c 842d8e8 783de55 17e971c 842d8e8 17e971c 842d8e8 783de55 842d8e8 783de55 842d8e8 783de55 842d8e8 783de55 842d8e8 783de55 842d8e8 783de55 842d8e8 783de55 842d8e8 783de55 842d8e8 783de55 842d8e8 783de55 842d8e8 17e971c 783de55 842d8e8 783de55 17e971c 842d8e8 17e971c 059c42b 17e971c 017d71a 17e971c | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 978 979 980 981 982 983 984 985 986 987 988 989 990 991 992 993 994 995 996 997 998 999 1000 1001 1002 1003 1004 1005 1006 1007 1008 1009 1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052 1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 1160 1161 1162 1163 1164 1165 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 1177 1178 1179 1180 1181 1182 1183 1184 1185 1186 1187 1188 1189 1190 1191 1192 1193 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 1205 1206 1207 1208 1209 1210 1211 1212 1213 1214 1215 1216 1217 1218 1219 1220 1221 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 1233 1234 1235 1236 1237 1238 1239 1240 1241 1242 1243 1244 1245 1246 1247 1248 1249 1250 1251 1252 1253 1254 1255 1256 1257 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 1294 1295 1296 1297 1298 1299 1300 1301 1302 1303 1304 1305 1306 1307 1308 1309 1310 1311 1312 1313 1314 1315 1316 1317 1318 1319 1320 1321 1322 1323 1324 1325 1326 1327 1328 1329 1330 1331 1332 1333 1334 1335 1336 1337 1338 1339 1340 1341 1342 1343 1344 1345 1346 1347 1348 1349 1350 1351 1352 1353 1354 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 1366 1367 1368 1369 1370 1371 1372 1373 1374 1375 1376 1377 1378 1379 1380 1381 1382 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 1394 1395 1396 1397 1398 1399 1400 1401 1402 1403 1404 1405 1406 1407 1408 1409 1410 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430 1431 1432 1433 1434 1435 1436 1437 1438 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518 1519 1520 1521 1522 1523 1524 1525 1526 1527 1528 1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553 1554 1555 1556 1557 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 1568 1569 1570 1571 1572 1573 1574 1575 1576 1577 1578 1579 1580 1581 1582 1583 1584 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 1596 1597 1598 1599 1600 1601 1602 1603 1604 1605 1606 1607 1608 1609 1610 1611 1612 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 1624 1625 1626 1627 1628 1629 1630 1631 1632 1633 1634 1635 1636 1637 1638 1639 1640 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 1652 1653 1654 1655 1656 1657 1658 1659 1660 1661 1662 1663 1664 1665 1666 1667 1668 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 1680 1681 1682 1683 1684 1685 1686 1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 1736 1737 1738 1739 1740 1741 1742 1743 1744 1745 1746 1747 1748 1749 1750 1751 1752 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 1764 1765 1766 1767 1768 1769 1770 1771 1772 1773 1774 1775 1776 1777 1778 1779 1780 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 1792 1793 1794 1795 1796 1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807 1808 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 1820 1821 1822 1823 1824 1825 1826 1827 1828 1829 1830 1831 1832 1833 1834 1835 1836 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 1848 1849 1850 1851 1852 1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 1876 1877 1878 1879 1880 1881 1882 1883 1884 1885 1886 1887 1888 1889 1890 1891 1892 1893 1894 1895 1896 1897 1898 1899 1900 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 1912 1913 1914 1915 1916 1917 1918 1919 1920 1921 1922 1923 1924 1925 1926 1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938 1939 1940 1941 1942 1943 1944 1945 1946 1947 1948 1949 1950 1951 1952 1953 1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 2036 2037 2038 2039 2040 2041 2042 2043 2044 2045 2046 2047 2048 2049 2050 2051 2052 2053 2054 2055 2056 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 2104 2105 2106 2107 2108 2109 2110 2111 2112 2113 2114 2115 2116 2117 2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 2139 2140 2141 2142 2143 2144 2145 2146 2147 2148 2149 2150 2151 2152 2153 2154 2155 2156 2157 2158 2159 2160 2161 2162 2163 2164 2165 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191 2192 2193 2194 2195 2196 2197 2198 2199 2200 2201 2202 2203 2204 2205 2206 2207 2208 2209 2210 2211 2212 2213 2214 2215 2216 2217 2218 2219 2220 2221 2222 2223 2224 2225 2226 2227 2228 2229 2230 2231 2232 2233 2234 2235 2236 2237 2238 2239 2240 2241 2242 2243 2244 2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286 2287 2288 2289 2290 2291 2292 2293 2294 2295 2296 2297 2298 2299 2300 2301 2302 2303 2304 2305 2306 2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327 2328 2329 2330 2331 2332 2333 2334 2335 2336 2337 2338 2339 2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355 2356 2357 2358 2359 2360 2361 2362 2363 2364 2365 2366 2367 2368 2369 2370 2371 2372 2373 2374 2375 2376 2377 2378 2379 2380 2381 2382 2383 2384 2385 2386 2387 2388 2389 2390 2391 2392 2393 2394 2395 2396 2397 2398 2399 2400 2401 2402 2403 2404 2405 2406 2407 2408 2409 2410 2411 2412 2413 2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431 2432 2433 2434 2435 2436 2437 2438 2439 2440 2441 2442 2443 2444 2445 2446 2447 2448 2449 2450 2451 2452 2453 2454 2455 2456 2457 2458 2459 2460 2461 2462 2463 2464 2465 2466 2467 2468 2469 2470 2471 2472 2473 2474 2475 2476 2477 2478 2479 2480 2481 2482 2483 2484 2485 2486 2487 2488 2489 2490 2491 2492 2493 2494 2495 2496 2497 2498 2499 2500 2501 2502 2503 2504 2505 2506 2507 2508 2509 2510 2511 2512 2513 2514 2515 2516 2517 2518 2519 2520 2521 2522 2523 2524 2525 2526 2527 2528 2529 2530 2531 2532 2533 2534 2535 2536 2537 2538 2539 2540 2541 2542 2543 2544 2545 2546 2547 2548 2549 2550 2551 2552 2553 2554 2555 2556 2557 2558 2559 2560 2561 2562 2563 2564 2565 2566 2567 2568 2569 2570 2571 2572 2573 2574 2575 2576 2577 2578 2579 2580 2581 2582 2583 2584 2585 2586 2587 2588 2589 2590 2591 2592 2593 2594 2595 2596 2597 2598 2599 2600 2601 2602 2603 2604 2605 2606 2607 2608 2609 2610 2611 2612 2613 2614 2615 2616 2617 2618 2619 2620 2621 2622 2623 2624 2625 2626 2627 2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661 2662 2663 2664 2665 2666 2667 2668 2669 2670 2671 2672 2673 2674 2675 2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687 2688 2689 2690 2691 2692 2693 2694 2695 2696 2697 2698 2699 2700 2701 2702 2703 2704 2705 2706 2707 2708 2709 2710 2711 2712 2713 2714 2715 2716 2717 2718 2719 2720 2721 2722 2723 2724 2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741 2742 2743 2744 2745 2746 2747 2748 2749 2750 2751 2752 2753 2754 2755 2756 2757 2758 2759 2760 2761 2762 2763 2764 2765 2766 2767 2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 2795 2796 2797 2798 2799 2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810 2811 2812 2813 2814 2815 2816 2817 2818 2819 2820 2821 2822 2823 2824 2825 2826 2827 2828 2829 2830 2831 2832 2833 2834 2835 2836 2837 2838 2839 2840 2841 2842 2843 2844 2845 2846 2847 2848 2849 2850 2851 2852 2853 2854 2855 2856 2857 2858 2859 2860 2861 2862 2863 2864 2865 2866 2867 2868 2869 2870 2871 2872 2873 2874 2875 2876 2877 2878 2879 2880 2881 2882 2883 2884 2885 2886 2887 2888 2889 2890 2891 2892 2893 2894 2895 2896 2897 2898 2899 2900 2901 2902 2903 2904 2905 2906 2907 2908 2909 2910 2911 2912 2913 2914 2915 2916 2917 2918 2919 2920 2921 2922 2923 2924 2925 2926 2927 2928 2929 2930 2931 2932 2933 2934 2935 2936 2937 2938 2939 2940 2941 2942 2943 2944 2945 2946 2947 2948 2949 2950 2951 2952 2953 2954 2955 2956 2957 2958 2959 2960 2961 2962 2963 2964 2965 2966 2967 2968 2969 2970 2971 2972 2973 2974 2975 2976 2977 2978 2979 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993 2994 2995 2996 2997 2998 2999 3000 3001 3002 3003 3004 3005 3006 3007 3008 3009 3010 3011 3012 3013 3014 3015 3016 3017 3018 3019 3020 3021 3022 3023 3024 3025 3026 3027 3028 3029 | #!/usr/bin/env python3
"""Backup and restore OpenClaw state to/from Hugging Face Dataset."""
from __future__ import annotations
import argparse
import datetime as dt
import fcntl
import hashlib
import json
import os
import re
import shutil
import subprocess
import sys
import tarfile
import tempfile
import time
import urllib.request
import uuid
from dataclasses import dataclass, field
from pathlib import Path
from typing import Any, Optional
def load_env_files():
"""Load environment variables from OpenClaw env files.
Note: HF Space variables (already in os.environ) take priority over env files.
This function only fills in missing variables from env files.
"""
env_files = [
"/etc/profile.d/openclaw-env.sh",
"/root/.env.d/openclaw-backup.env",
]
for env_file in env_files:
if os.path.isfile(env_file):
try:
result = subprocess.run(
["bash", "-c", f"source '{env_file}' 2>/dev/null && env"],
capture_output=True,
text=True,
timeout=10,
)
if result.returncode == 0:
for line in result.stdout.strip().split("\n"):
if "=" in line:
key, _, value = line.partition("=")
if key and key not in os.environ:
os.environ[key] = value
except Exception:
pass
def env_get_with_hf_priority(name: str, default: str = "") -> str:
"""Get environment variable with HF Space variable taking priority.
Priority:
1. Existing os.environ value (HF Space variable)
2. /etc/profile.d/openclaw-env.sh
3. /root/.env.d/openclaw-backup.env
4. default value
Args:
name: Environment variable name
default: Default value if not found anywhere
Returns:
The environment variable value
"""
if name in os.environ and os.environ[name]:
return os.environ[name]
env_files = [
"/etc/profile.d/openclaw-env.sh",
"/root/.env.d/openclaw-backup.env",
]
for env_file in env_files:
if os.path.isfile(env_file):
try:
result = subprocess.run(
["bash", "-c", f"source '{env_file}' 2>/dev/null && echo ${name}"],
capture_output=True,
text=True,
timeout=10,
)
if result.returncode == 0:
value = result.stdout.strip()
if value:
return value
except Exception:
pass
return default
# Set Hugging Face Hub timeout before importing (in seconds)
# This helps avoid timeout errors when downloading large backup files
os.environ.setdefault("HF_HUB_DOWNLOAD_TIMEOUT", "300") # 5 minutes
os.environ.setdefault("HF_HUB_ETAG_TIMEOUT", "60") # 60 seconds timeout for metadata
try:
from huggingface_hub import HfApi, hf_hub_download
from huggingface_hub.utils import HfHubHTTPError
except ModuleNotFoundError: # pragma: no cover - runtime dependency, optional for unit tests
HfApi = None # type: ignore[assignment]
hf_hub_download = None # type: ignore[assignment]
class HfHubHTTPError(Exception): # type: ignore[no-redef]
"""Fallback error when huggingface_hub is not installed."""
def env_bool(name: str, default: bool) -> bool:
raw = os.getenv(name)
if raw is None:
return default
return raw.strip().lower() in {"1", "true", "yes", "on"}
def utc_now() -> dt.datetime:
return dt.datetime.now(dt.UTC)
# 元数据版本常量
METADATA_VERSION = "2.1" # 当前元数据版本
MIN_SUPPORTED_METADATA_VERSION = "1.0" # 最低支持的元数据版本
def parse_metadata_version(version_str: str) -> tuple[int, int]:
"""解析元数据版本字符串为版本号元组。"""
try:
parts = version_str.split(".")
major = int(parts[0])
minor = int(parts[1]) if len(parts) > 1 else 0
return (major, minor)
except (ValueError, IndexError):
return (1, 0) # 默认返回1.0版本
def is_metadata_compatible(version_str: str) -> bool:
"""检查元数据版本是否兼容当前版本。
兼容性规则：
- 目标主版本 <= 当前主版本：向后兼容（旧备份可以用新版本恢复）
- 目标主版本 > 当前主版本：不兼容（新备份不能用旧版本恢复，需升级）
- 次版本在支持范围内
"""
current = parse_metadata_version(METADATA_VERSION)
target = parse_metadata_version(version_str)
min_supported = parse_metadata_version(MIN_SUPPORTED_METADATA_VERSION)
# 目标主版本大于当前主版本：不兼容（需要升级软件）
if target[0] > current[0]:
return False
# 目标主版本小于当前主版本：向后兼容（旧备份可用新版本恢复）
# 但仍需检查次版本是否在支持范围内
if target[0] < current[0]:
return min_supported[1] <= target[1]
# 主版本相同：次版本必须在支持范围内
return min_supported[1] <= target[1] <= current[1]
@dataclass
class BackupConfig:
dataset_repo: str
state_dir: Path
backup_source_dir: Path
work_dir: Path
metadata_dir: Path # Directory for backup metadata (last-backup-metadata.json)
repo_type: str
path_prefix: str
private: bool
# Multi-dataset configuration for flexible restore/backup scenarios
restore_dataset_repo: Optional[str] = None # Dataset to restore from (defaults to dataset_repo)
backup_npm_enabled: bool = True # Whether to backup npm directory
restore_npm_enabled: bool = True # Whether to restore npm directory
# Note: Container restart/rebuild always triggers restore. This is by design.
incremental: bool = True # 默认开启增量备份
incremental_interval_minutes: int = 10 # 增量备份间隔（分钟）
compression_level: int = 6
split_size: str = "" # e.g., "500M" for 500MB volumes
size_warning_mb: int = 1500 # Warn if backup exceeds this size
# Dynamic backup strategy settings
dynamic_backup: bool = True # 默认开启动态备份策略
dynamic_small_threshold_mb: int = 500
dynamic_medium_threshold_mb: int = 2000
dynamic_high_change_rate: int = 10
dynamic_low_change_rate: int = 2
dynamic_min_changed_files: int = 5 # Minimum changed files to consider backup worthwhile
dynamic_min_changed_size_kb: int = 100 # Minimum total size of changed files (KB)
# Full backup strategy settings
full_backup_interval_hours: int = 2 # Force full backup after this many hours
max_incremental_backups: int = 10 # Force full backup after this many incremental backups
root_config_dir: Path = Path("/root/.config").resolve()
root_codex_dir: Path = Path("/root/.codex").resolve()
root_claude_dir: Path = Path("/root/.claude").resolve()
root_agents_dir: Path = Path("/root/.agents").resolve()
root_ssh_dir: Path = Path("/root/.ssh").resolve()
root_env_dir: Path = Path("/root/.env.d").resolve()
root_npm_dir: Path = Path("/root/.npm").resolve()
root_lark_cli_dir: Path = Path("/root/.lark-cli").resolve()
extra_dirs: dict[str, Path] = field(default_factory=dict)
extra_files: dict[str, Path] = field(default_factory=dict)
keep_count: int = 48
# Local backup retention settings
keep_local_backup: bool = False # Whether to keep a local copy after upload
local_backup_dir: Optional[Path] = None # Directory to store local backups (None = use work_dir)
keep_local_count: int = 5 # Number of local backups to keep
# Encryption settings
encryption_enabled: bool = False # Whether to encrypt backups
encryption_password: Optional[str] = None # Password for encryption (from HF secret)
@classmethod
def from_env(cls) -> "BackupConfig":
# Load environment from files first (fills in missing vars only)
load_env_files()
# HF Space variables take priority over env files
dataset_repo = env_get_with_hf_priority("OPENCLAW_BACKUP_DATASET_REPO")
if not dataset_repo:
raise ValueError("OPENCLAW_BACKUP_DATASET_REPO is not set")
# Multi-dataset configuration
# HF Space variable takes priority, fallback to dataset_repo
restore_dataset_repo = env_get_with_hf_priority("OPENCLAW_RESTORE_DATASET_REPO", dataset_repo)
backup_npm_enabled = env_get_with_hf_priority("OPENCLAW_BACKUP_NPM_ENABLED", "true").lower() in ("true", "1", "yes")
restore_npm_enabled = env_get_with_hf_priority("OPENCLAW_RESTORE_NPM_ENABLED", "true").lower() in ("true", "1", "yes")
# Note: Container restart/rebuild always triggers restore. This is by design.
# Encryption configuration
encryption_enabled = env_bool("OPENCLAW_BACKUP_ENCRYPTION_ENABLED", False)
encryption_password = os.getenv("OPENCLAW_BACKUP_ENCRYPTION_PASSWORD") or None
state_dir = Path(os.getenv("OPENCLAW_STATE_DIR", "/root/.openclaw")).resolve()
backup_source_raw = (os.getenv("OPENCLAW_BACKUP_SOURCE_DIR") or "").strip()
if backup_source_raw:
backup_source_dir = Path(backup_source_raw).resolve()
else:
backup_source_dir = state_dir
root_config_dir = Path(
(os.getenv("OPENCLAW_BACKUP_ROOT_CONFIG_DIR") or "/root/.config").strip() or "/root/.config"
).resolve()
root_codex_dir = Path(
(os.getenv("OPENCLAW_BACKUP_ROOT_CODEX_DIR") or "/root/.codex").strip() or "/root/.codex"
).resolve()
root_claude_dir = Path(
(os.getenv("OPENCLAW_BACKUP_ROOT_CLAUDE_DIR") or "/root/.claude").strip() or "/root/.claude"
).resolve()
root_agents_dir = Path(
(os.getenv("OPENCLAW_BACKUP_ROOT_AGENTS_DIR") or "/root/.agents").strip() or "/root/.agents"
).resolve()
root_ssh_dir = Path((os.getenv("OPENCLAW_BACKUP_ROOT_SSH_DIR") or "/root/.ssh").strip() or "/root/.ssh").resolve()
root_env_dir = Path((os.getenv("OPENCLAW_BACKUP_ROOT_ENV_DIR") or "/root/.env.d").strip() or "/root/.env.d").resolve()
root_npm_dir = Path((os.getenv("OPENCLAW_BACKUP_ROOT_NPM_DIR") or "/root/.npm").strip() or "/root/.npm").resolve()
root_lark_cli_dir = Path(
(os.getenv("OPENCLAW_BACKUP_ROOT_LARK_CLI_DIR") or "/root/.lark-cli").strip() or "/root/.lark-cli"
).resolve()
work_dir = Path(os.getenv("OPENCLAW_BACKUP_WORK_DIR", "/tmp/openclaw-backup")).resolve()
metadata_dir = Path(os.getenv("OPENCLAW_BACKUP_METADATA_DIR", "/root/.backup-info")).resolve()
repo_type = (os.getenv("OPENCLAW_BACKUP_REPO_TYPE") or "dataset").strip() or "dataset"
path_prefix = (os.getenv("OPENCLAW_BACKUP_PATH_PREFIX") or "backups").strip("/")
private = env_bool("OPENCLAW_BACKUP_PRIVATE", True)
keep_count = 48
keep_count_raw = (os.getenv("OPENCLAW_BACKUP_KEEP_COUNT") or "").strip()
if keep_count_raw:
try:
parsed_keep_count = int(keep_count_raw)
if parsed_keep_count >= 1:
keep_count = parsed_keep_count
else:
print(
f"invalid OPENCLAW_BACKUP_KEEP_COUNT={keep_count_raw!r} (must be >= 1), using default 48",
file=sys.stderr,
)
except ValueError:
print(
f"invalid OPENCLAW_BACKUP_KEEP_COUNT={keep_count_raw!r} (not a number), using default 48",
file=sys.stderr,
)
# Parse extra directories from OPENCLAW_BACKUP_EXTRA_DIRS
# Format: "archive_name:/path/to/dir,archive_name2:/path/to/dir2"
extra_dirs: dict[str, Path] = {}
extra_dirs_raw = (os.getenv("OPENCLAW_BACKUP_EXTRA_DIRS") or "").strip()
if extra_dirs_raw:
for item in extra_dirs_raw.split(","):
item = item.strip()
if not item:
continue
if ":" not in item:
print(
f"invalid extra dir entry (missing ':'): {item!r}, skipping",
file=sys.stderr,
)
continue
archive_name, dir_path = item.split(":", 1)
archive_name = archive_name.strip()
dir_path = dir_path.strip()
if not archive_name or not dir_path:
print(
f"invalid extra dir entry (empty name or path): {item!r}, skipping",
file=sys.stderr,
)
continue
extra_dirs[archive_name] = Path(dir_path).resolve()
# Parse extra files from OPENCLAW_BACKUP_EXTRA_FILES
# Format: "archive_name:/path/to/file,archive_name2:/path/to/file2"
extra_files: dict[str, Path] = {}
extra_files_raw = (os.getenv("OPENCLAW_BACKUP_EXTRA_FILES") or "").strip()
if extra_files_raw:
for item in extra_files_raw.split(","):
item = item.strip()
if not item:
continue
if ":" not in item:
print(
f"invalid extra file entry (missing ':'): {item!r}, skipping",
file=sys.stderr,
)
continue
archive_name, file_path = item.split(":", 1)
archive_name = archive_name.strip()
file_path = file_path.strip()
if not archive_name or not file_path:
print(
f"invalid extra file entry (empty name or path): {item!r}, skipping",
file=sys.stderr,
)
continue
extra_files[archive_name] = Path(file_path).resolve()
return cls(
dataset_repo=dataset_repo,
state_dir=state_dir,
backup_source_dir=backup_source_dir,
work_dir=work_dir,
metadata_dir=metadata_dir,
repo_type=repo_type,
path_prefix=path_prefix,
private=private,
incremental=env_bool("OPENCLAW_INCREMENTAL_BACKUP", True),
incremental_interval_minutes=int(os.getenv("OPENCLAW_INCREMENTAL_INTERVAL_MINUTES", "5")),
compression_level=int(os.getenv("OPENCLAW_BACKUP_COMPRESSION_LEVEL", "6")),
split_size=(os.getenv("OPENCLAW_BACKUP_SPLIT_SIZE") or "").strip(),
size_warning_mb=int(os.getenv("OPENCLAW_BACKUP_SIZE_WARNING_MB", "1500")),
# Dynamic backup strategy
dynamic_backup=env_bool("OPENCLAW_DYNAMIC_BACKUP", True),
dynamic_small_threshold_mb=int(os.getenv("OPENCLAW_DYNAMIC_SMALL_THRESHOLD_MB", "500")),
dynamic_medium_threshold_mb=int(os.getenv("OPENCLAW_DYNAMIC_MEDIUM_THRESHOLD_MB", "2000")),
dynamic_high_change_rate=int(os.getenv("OPENCLAW_DYNAMIC_HIGH_CHANGE_RATE", "10")),
dynamic_low_change_rate=int(os.getenv("OPENCLAW_DYNAMIC_LOW_CHANGE_RATE", "2")),
dynamic_min_changed_files=int(os.getenv("OPENCLAW_DYNAMIC_MIN_CHANGED_FILES", "5")),
dynamic_min_changed_size_kb=int(os.getenv("OPENCLAW_DYNAMIC_MIN_CHANGED_SIZE_KB", "100")),
full_backup_interval_hours=int(os.getenv("OPENCLAW_FULL_BACKUP_INTERVAL_HOURS", "2")),
max_incremental_backups=int(os.getenv("OPENCLAW_MAX_INCREMENTAL_BACKUPS", "10")),
root_config_dir=root_config_dir,
root_codex_dir=root_codex_dir,
root_claude_dir=root_claude_dir,
root_agents_dir=root_agents_dir,
root_ssh_dir=root_ssh_dir,
root_env_dir=root_env_dir,
root_npm_dir=root_npm_dir,
root_lark_cli_dir=root_lark_cli_dir,
extra_dirs=extra_dirs,
extra_files=extra_files,
keep_count=keep_count,
# Multi-dataset configuration
restore_dataset_repo=restore_dataset_repo,
backup_npm_enabled=backup_npm_enabled,
restore_npm_enabled=restore_npm_enabled,
# Encryption settings
encryption_enabled=encryption_enabled,
encryption_password=encryption_password,
)
class OpenClawBackup:
_ARCHIVE_NAME_RE = re.compile(r"^openclaw-backup-(\d{8}-\d{6})\.tar\.gz(?:\.enc)?(?:\.part-[a-z]{2})?$")
def __init__(self, config: BackupConfig, api: Optional[Any] = None, token: Optional[str] = None) -> None:
self.config = config
self.token = self._resolve_token(token)
config.metadata_dir.mkdir(parents=True, exist_ok=True)
self._meta_cache: dict[str, dict] = {}
self._meta_file_cache: dict[str, Path] = {}
if api is not None:
self.api = api
return
if HfApi is None:
raise ModuleNotFoundError(
"huggingface_hub is required. Install with: pip install 'huggingface_hub[cli]'"
)
self.api = HfApi(token=self.token)
def _resolve_token(self, explicit_token: Optional[str] = None) -> Optional[str]:
"""Resolve HF token with proper fallback hierarchy.
Priority:
1. Explicitly passed token (from constructor)
2. HF_TOKEN from os.environ (HF Space variables, highest priority)
3. HUGGINGFACE_HUB_TOKEN from os.environ
4. HF_TOKEN from env files (/etc/profile.d/openclaw-env.sh, /root/.env.d/openclaw-backup.env)
5. HUGGINGFACE_HUB_TOKEN from env files
6. None (no token available)
"""
if explicit_token:
return explicit_token
for env_var in ("HF_TOKEN", "HUGGINGFACE_HUB_TOKEN"):
token = os.getenv(env_var)
if token:
return token
env_files = [
"/etc/profile.d/openclaw-env.sh",
"/root/.env.d/openclaw-backup.env",
]
for env_file in env_files:
if os.path.isfile(env_file):
try:
result = subprocess.run(
["bash", "-c", f"source '{env_file}' 2>/dev/null && env"],
capture_output=True,
text=True,
timeout=10,
)
if result.returncode == 0:
for line in result.stdout.strip().split("\n"):
if "=" in line:
key, _, value = line.partition("=")
if key in ("HF_TOKEN", "HUGGINGFACE_HUB_TOKEN") and value:
return value
except Exception:
pass
return None
@staticmethod
def _calculate_checksum(file_path: Path, algorithm: str = "sha256") -> str:
"""Calculate checksum for a file.
Enterprise-grade data integrity verification.
"""
hash_obj = hashlib.new(algorithm)
with open(file_path, "rb") as f:
for chunk in iter(lambda: f.read(8192), b""):
hash_obj.update(chunk)
return hash_obj.hexdigest()
def _verify_backup_integrity(self, archive_path: Path, is_split: bool = False) -> dict:
"""Verify backup archive integrity.
For split volumes, only checks existence and size (cannot validate tar structure of partial file).
Returns a dictionary with verification results.
"""
result = {
"valid": False,
"checksum": None,
"size": 0,
"error": None,
}
try:
# Check file exists and is readable
if not archive_path.exists():
result["error"] = "Archive does not exist"
return result
# Check file size
result["size"] = archive_path.stat().st_size
if result["size"] == 0:
result["error"] = "Archive is empty"
return result
# Calculate checksum
result["checksum"] = self._calculate_checksum(archive_path)
# For split volumes, skip tar validation (partial file)
if not is_split:
# Verify tar archive can be opened
with tarfile.open(archive_path, "r:gz") as tar:
# Try to list members (this validates the archive structure)
members = tar.getmembers()
if not members:
result["error"] = "Archive is empty (no members)"
return result
# Debug: print first few members
print(f" Archive contains {len(members)} entries")
for m in members[:10]:
print(f" - {m.name}")
if len(members) > 10:
print(f" ... and {len(members) - 10} more entries")
# Archive structure validation passed
# Note: valid archive roots include predefined ones (openclaw-state,
# root-config, etc.) plus custom ones from extra_dirs/extra_files env vars.
# Since custom names are user-defined, we only check that:
# 1. Archive is a valid tar file (proven by being able to list members)
# 2. Archive is not empty (checked above)
# The specific root directory names are validated by the backup source
# configuration, not here.
result["valid"] = True
except tarfile.TarError as e:
result["error"] = f"Invalid tar archive: {e}"
except Exception as e:
result["error"] = f"Verification failed: {e}"
return result
def ensure_dataset_repo(self) -> None:
"""Ensure the backup target repository exists.
"""
target_repo = self.config.dataset_repo
self.api.create_repo(
repo_id=target_repo,
repo_type=self.config.repo_type,
private=self.config.private,
exist_ok=True,
)
def _check_repo_exists(self, repo_id: str, repo_type: str = "dataset") -> bool:
"""Check if a repository exists.
Args:
repo_id: Repository ID to check
repo_type: Repository type (dataset, model, space)
Returns:
True if repository exists, False otherwise
"""
try:
self.api.list_repo_files(
repo_id=repo_id,
repo_type=repo_type,
)
return True
except Exception as e:
error_str = str(e).lower()
if "not found" in error_str or "404" in error_str or "does not exist" in error_str:
return False
raise
def _get_restore_repo(self) -> str:
"""Get the repository to restore from, with fallback logic.
If OPENCLAW_RESTORE_DATASET_REPO is set but doesn't exist,
falls back to OPENCLAW_BACKUP_DATASET_REPO.
Returns:
Repository ID to restore from
"""
restore_repo = self.config.restore_dataset_repo
backup_repo = self.config.dataset_repo
if restore_repo and restore_repo != backup_repo:
if self._check_repo_exists(restore_repo, self.config.repo_type):
print(f"Using restore dataset: {restore_repo}")
return restore_repo
else:
print(f"Warning: Restore dataset '{restore_repo}' does not exist or is not accessible")
print(f"Falling back to backup dataset: {backup_repo}")
return backup_repo
print(f"Using backup dataset for restore: {backup_repo}")
return backup_repo
def _calculate_change_rate(self) -> float:
"""Calculate file change rate (files per minute) since last backup."""
rate, _ = self._calculate_change_rate_and_count()
return rate
def _calculate_change_rate_and_count(self) -> tuple[float, int]:
"""Calculate file change rate (files per minute) and count since last backup.
Returns:
Tuple of (change_rate, changed_files_count)
"""
last_backup_time = self._get_last_backup_time()
if last_backup_time is None:
return (float('inf'), 0) # First backup, treat as high change rate
minutes_since_last = (utc_now() - last_backup_time).total_seconds() / 60
if minutes_since_last < 1:
minutes_since_last = 1 # Avoid division by zero
# Count recently modified files
changed_files = 0
for directory in [self.config.backup_source_dir] + list(self._extra_root_dirs().values()):
if directory.exists():
changed_files += len(self._get_changed_files(directory, last_backup_time))
return (changed_files / minutes_since_last, changed_files)
def _estimate_backup_size_mb(self) -> float:
"""Estimate the size of the next backup in MB."""
total_size = 0
if not self.config.incremental or self._should_do_full_backup(use_cached=True):
# Full backup - calculate total size
for directory in [self.config.backup_source_dir] + list(self._extra_root_dirs().values()):
if directory.exists():
for item in directory.rglob("*"):
if item.is_file():
try:
total_size += item.stat().st_size
except OSError:
continue
else:
# Incremental backup - only changed files
last_backup_time = self._get_last_backup_time()
if last_backup_time:
for directory in [self.config.backup_source_dir] + list(self._extra_root_dirs().values()):
if directory.exists():
for file_path in self._get_changed_files(directory, last_backup_time):
try:
total_size += file_path.stat().st_size
except OSError:
continue
return total_size / (1024 * 1024)
def _apply_dynamic_strategy(self, force_full_backup: bool = False) -> dict[str, any]:
"""Apply dynamic backup strategy based on file size and change rate.
Args:
force_full_backup: If True, skip checking for incremental backup eligibility
(backup type decision already made by caller).
"""
if not self.config.dynamic_backup:
return {
"compression_level": self.config.compression_level,
"split_size": self.config.split_size,
"skip_backup": False,
"reason": "Dynamic backup disabled",
}
# Get current metrics
estimated_size_mb = self._estimate_backup_size_mb()
estimated_size_kb = estimated_size_mb * 1024
change_rate, changed_files = self._calculate_change_rate_and_count()
print(f"[Dynamic Strategy] Estimated size: {estimated_size_mb:.1f}MB, Change rate: {change_rate:.1f} files/min, Changed files: {changed_files}")
# Determine file size category
if estimated_size_mb < self.config.dynamic_small_threshold_mb:
size_category = "small"
compression = 3 # Fast compression for small files
split = ""
elif estimated_size_mb < self.config.dynamic_medium_threshold_mb:
size_category = "medium"
compression = 6 # Balanced for medium files
split = "500M" if estimated_size_mb > 1000 else ""
else:
size_category = "large"
compression = 9 # Maximum compression for large files
split = "500M"
# Adjust based on change rate
skip_reason = None
if change_rate > self.config.dynamic_high_change_rate:
# High change rate - prioritize speed over compression
compression = max(1, compression - 2)
elif change_rate < self.config.dynamic_low_change_rate:
# Low change rate - can use higher compression
compression = min(9, compression + 1)
# Skip if very low change rate AND (fewer changed files OR smaller total size) than minimum threshold
# But still perform backup if it's a scheduled full backup
if changed_files < self.config.dynamic_min_changed_files and estimated_size_kb < self.config.dynamic_min_changed_size_kb and not force_full_backup:
current_skipped = self._load_skipped_count() + 1
self._save_skipped_count(current_skipped)
max_skipped_incrementals = max(3, self.config.max_incremental_backups // 3)
if current_skipped >= max_skipped_incrementals:
print(f"[Dynamic Strategy] Too many skipped incrementals ({current_skipped}), forcing backup to maintain chain continuity")
self._clear_skipped_count()
else:
skip_reason = f"Very low change rate ({change_rate:.1f} files/min, {changed_files} files, {estimated_size_kb:.0f}KB), skipping incremental backup ({current_skipped}/{max_skipped_incrementals})"
print(f"[Dynamic Strategy] {skip_reason}")
return {
"compression_level": compression,
"split_size": split,
"skip_backup": True,
"reason": skip_reason,
}
else:
self._clear_skipped_count()
# Adjust compression based on size category
if size_category == "large" and estimated_size_mb > 3000:
# Very large files - always use max compression
compression = 9
strategy = {
"compression_level": compression,
"split_size": split,
"skip_backup": False,
"reason": f"Size: {size_category} ({estimated_size_mb:.1f}MB), Change rate: {change_rate:.1f} files/min",
}
print(f"[Dynamic Strategy] Applied: compression={compression}, split={split or 'none'}")
return strategy
def _extra_root_dirs(self) -> dict[str, Path]:
dirs = {
"root-config": self.config.root_config_dir,
"root-codex": self.config.root_codex_dir,
"root-claude": self.config.root_claude_dir,
"root-agents": self.config.root_agents_dir,
"root-ssh": self.config.root_ssh_dir,
"root-env": self.config.root_env_dir,
"root-lark-cli": self.config.root_lark_cli_dir,
"backup-info": self.config.metadata_dir,
}
if self.config.backup_npm_enabled:
dirs["root-npm"] = self.config.root_npm_dir
if self.config.extra_dirs:
dirs.update(self.config.extra_dirs)
return dirs
def _restore_root_dirs(self) -> dict[str, Path]:
dirs = {
"root-config": self.config.root_config_dir,
"root-codex": self.config.root_codex_dir,
"root-claude": self.config.root_claude_dir,
"root-agents": self.config.root_agents_dir,
"root-ssh": self.config.root_ssh_dir,
"root-env": self.config.root_env_dir,
"root-lark-cli": self.config.root_lark_cli_dir,
"backup-info": self.config.metadata_dir,
}
if self.config.restore_npm_enabled:
dirs["root-npm"] = self.config.root_npm_dir
if self.config.extra_dirs:
dirs.update(self.config.extra_dirs)
return dirs
def _get_last_backup_time(self) -> Optional[dt.datetime]:
"""Get the timestamp of the last successful backup from metadata.
Checks metadata_dir for last-backup-metadata.json, then falls back to
downloading latest-backup.json and the corresponding .meta.json from
remote storage.
"""
metadata_path = self.config.metadata_dir / "last-backup-metadata.json"
try:
if metadata_path.exists():
with open(metadata_path, "r") as f:
metadata = json.load(f)
last_backup_str = metadata.get("last_backup_time")
if last_backup_str:
return dt.datetime.fromisoformat(last_backup_str)
except Exception:
pass
return self._get_last_backup_time_from_remote()
def _get_remote_backup_chain_info(self) -> Optional[dict]:
"""Get the complete backup chain info from remote storage.
Downloads latest-backup.json and the latest .meta.json to get:
- last_backup_time: timestamp of last backup
- parent_meta_path: path to parent backup's .meta.json
- chain_id: the chain this backup belongs to
- volumes: list of volume filenames (for split backups)
- backup_type: "full" or "incremental"
This is the source of truth for backup chain continuity.
"""
try:
latest_json_path = "latest-backup.json"
latest_json_local = self._download_backup(latest_json_path, self.config.work_dir)
with open(latest_json_local, "r") as f:
latest_json = json.load(f)
meta_path = latest_json.get("latest", "")
if not meta_path:
return None
meta_local = self._download_backup(meta_path, self.config.work_dir)
with open(meta_local, "r") as f:
meta = json.load(f)
meta_filename = Path(meta_path).name
base_archive_name = meta_filename.replace(".meta.json", "")
return {
"last_backup_time": meta.get("last_backup_time") or meta.get("created_at_utc"),
"parent_meta_path": meta.get("parent"), # The parent field from the .meta.json file
"chain_id": meta.get("chain_id"),
"volumes": meta.get("volumes", [base_archive_name]),
"backup_type": meta.get("backup_type", "incremental"),
"is_split": meta.get("is_split", False),
}
except Exception:
return None
def _get_last_backup_time_from_remote(self) -> Optional[dt.datetime]:
"""Download latest-backup.json and corresponding .meta.json from remote to get last backup time.
This is used when local metadata is not available (e.g., first run on a new machine).
"""
try:
latest_json_path = "latest-backup.json"
print(f"[_get_last_backup_time_from_remote] Downloading {latest_json_path}...")
latest_json_local = self._download_backup(latest_json_path, self.config.work_dir)
print(f"[_get_last_backup_time_from_remote] Downloaded to {latest_json_local}")
with open(latest_json_local, "r") as f:
latest_json = json.load(f)
meta_path = latest_json.get("latest", "")
print(f"[_get_last_backup_time_from_remote] latest path: {meta_path}")
if not meta_path:
return None
print(f"[_get_last_backup_time_from_remote] Downloading {meta_path}...")
cached_meta = self._get_cached_meta(meta_path)
if cached_meta is not None:
meta = cached_meta
print(f"[_get_last_backup_time_from_remote] Using cached metadata for {meta_path}")
else:
meta_local = self._download_backup(meta_path, self.config.work_dir)
print(f"[_get_last_backup_time_from_remote] Downloaded to {meta_local}")
with open(meta_local, "r") as f:
meta = json.load(f)
self._set_cached_meta(meta_path, meta)
last_backup_str = meta.get("last_backup_time")
print(f"[_get_last_backup_time_from_remote] last_backup_time: {last_backup_str}")
if last_backup_str:
return dt.datetime.fromisoformat(last_backup_str)
except Exception as e:
print(f"[_get_last_backup_time_from_remote] Error: {type(e).__name__}: {e}")
pass
return None
def _save_backup_metadata(
self,
archive_path: Path,
file_count: int,
is_full_backup: bool = True,
parent_meta_path: Optional[str] = None,
chain_id: Optional[str] = None
) -> None:
"""Save metadata about this backup for future incremental backups.
This creates the per-backup .meta.json file content.
The metadata is saved locally and also uploaded to remote storage.
Args:
archive_path: Path to the archive file (may be split into multiple volumes)
file_count: Number of files in the backup
is_full_backup: Whether this is a full backup
parent_meta_path: Path to parent backup's .meta.json file (for incremental)
chain_id: Chain ID for this backup (generated for full, inherited for incremental)
"""
if chain_id is None:
chain_id = str(uuid.uuid4())[:8]
prev_metadata = self._get_backup_metadata()
if is_full_backup:
backup_type = "full"
chain_id = str(uuid.uuid4())[:8] # Full backup generates new chain_id
else:
backup_type = "incremental"
# For incremental, use provided chain_id or inherit from remote
if not chain_id:
chain_id = prev_metadata.get("chain_id", str(uuid.uuid4())[:8])
checksum = self._calculate_checksum(archive_path)
archive_size = archive_path.stat().st_size
if is_full_backup:
incremental_count = 0
last_full_backup_time = utc_now().isoformat(timespec="seconds")
else:
incremental_count = (prev_metadata.get("incremental_count", 0) + 1) if prev_metadata else 1
last_full_backup_time = prev_metadata.get("last_full_backup_time") if prev_metadata else None
metadata = {
"version": METADATA_VERSION,
"backup_type": backup_type,
"chain_id": chain_id,
"parent": parent_meta_path,
"volumes": [archive_path.name],
"checksum": f"sha256:{checksum}",
"created_at_utc": utc_now().isoformat(timespec="seconds"),
"last_backup_time": utc_now().isoformat(timespec="seconds"),
"file_count": file_count,
"archive_size": archive_size,
"is_latest": True,
"created_by": "openclaw-backup",
"incremental_count": incremental_count,
"last_full_backup_time": last_full_backup_time,
}
if prev_metadata:
prev_metadata["is_latest"] = False
prev_path = self.config.metadata_dir / "last-backup-metadata.json"
with open(prev_path, "w") as f:
json.dump(prev_metadata, f, indent=2)
metadata_path = self.config.metadata_dir / "last-backup-metadata.json"
with open(metadata_path, "w") as f:
json.dump(metadata, f, indent=2)
def _get_backup_metadata(self) -> Optional[dict]:
"""Get the current backup metadata from local files.
Checks metadata_dir for last-backup-metadata.json.
"""
metadata_path = self.config.metadata_dir / "last-backup-metadata.json"
try:
if metadata_path.exists():
with open(metadata_path, "r") as f:
return json.load(f)
except Exception:
pass
return None
def _get_skipped_count_file(self) -> Path:
"""Get path to the skipped incremental count file."""
return self.config.metadata_dir / ".skipped-incremental-count"
def _load_skipped_count(self) -> int:
"""Load the count of consecutively skipped incremental backups."""
skipped_file = self._get_skipped_count_file()
try:
if skipped_file.exists():
return int(skipped_file.read_text().strip())
except (ValueError, OSError):
pass
return 0
def _save_skipped_count(self, count: int) -> None:
"""Save the count of consecutively skipped incremental backups."""
skipped_file = self._get_skipped_count_file()
try:
self.config.metadata_dir.mkdir(parents=True, exist_ok=True)
skipped_file.write_text(str(count))
except OSError:
pass
def _clear_skipped_count(self) -> None:
"""Clear the skipped incremental count (when a backup is performed)."""
skipped_file = self._get_skipped_count_file()
try:
skipped_file.unlink(missing_ok=True)
except OSError:
pass
def _update_volumes_in_metadata(self, volumes: list[str], checksum: Optional[str] = None, archive_size: Optional[int] = None) -> None:
"""Update volumes field in metadata after split/encryption.
This is necessary because _save_backup_metadata is called BEFORE split/encryption,
so it records the original archive name. After splitting and/or encrypting, we need
to update with the actual volume names and checksums.
Args:
volumes: List of split volume filenames
checksum: New checksum for the encrypted archive (optional)
archive_size: New archive size in bytes (optional)
"""
try:
metadata_path = self.config.metadata_dir / "last-backup-metadata.json"
if metadata_path.exists():
with open(metadata_path, "r") as f:
metadata = json.load(f)
metadata["volumes"] = volumes
if checksum:
metadata["checksum"] = checksum
if archive_size is not None:
metadata["archive_size"] = archive_size
if any(v.endswith(".enc") for v in volumes):
metadata["encrypted"] = True
metadata["encryption_algorithm"] = "AES-256-CBC"
if any(".part-" in v for v in volumes):
metadata["is_split"] = True
with open(metadata_path, "w") as f:
json.dump(metadata, f, indent=2)
except Exception as e:
print(f"Warning: Could not update metadata: {e}")
def _should_do_full_backup(self, use_cached: bool = False) -> bool:
"""Determine if a full backup should be performed instead of incremental.
Args:
use_cached: If True, return cached result from previous call in same backup cycle.
If False, always compute fresh (but will cache if this is first call).
"""
cached_attr = '_cached_should_full_backup'
if use_cached and hasattr(self, cached_attr):
return getattr(self, cached_attr)
if not self.config.incremental:
return True
last_backup_time = self._get_last_backup_time()
if last_backup_time is None:
print("[DEBUG] No previous backup found, performing full backup")
result = True
else:
time_since_last = utc_now() - last_backup_time
if time_since_last.total_seconds() < self.config.incremental_interval_minutes * 60:
print(f"[DEBUG] Incremental interval not reached ({time_since_last.total_seconds():.0f}s < {self.config.incremental_interval_minutes * 60}s), skipping backup")
result = False
else:
result = self._should_force_full_backup()
setattr(self, cached_attr, result)
return result
def _should_force_full_backup(self) -> bool:
"""Check if full backup should be forced based on interval/count rules."""
metadata = self._get_backup_metadata()
last_full_backup_time_str = None
incremental_count = 0
if metadata:
last_full_backup_time_str = metadata.get("last_full_backup_time")
incremental_count = metadata.get("incremental_count", 0)
if last_full_backup_time_str:
last_full_backup_time = dt.datetime.fromisoformat(last_full_backup_time_str)
hours_since_full = (utc_now() - last_full_backup_time).total_seconds() / 3600
if hours_since_full >= self.config.full_backup_interval_hours:
print(f"[DEBUG] Full backup interval reached ({hours_since_full:.1f}h >= {self.config.full_backup_interval_hours}h), forcing full backup")
return True
if incremental_count >= self.config.max_incremental_backups:
print(f"[DEBUG] Max incremental backups reached ({incremental_count}/{self.config.max_incremental_backups}), forcing full backup")
return True
return False
def _get_changed_files(self, directory: Path, since: dt.datetime) -> list[Path]:
"""Get list of files that have been modified since the given timestamp."""
changed_files = []
if not directory.exists():
return changed_files
for item in directory.rglob("*"):
if item.is_file():
try:
mtime = dt.datetime.fromtimestamp(item.stat().st_mtime, tz=dt.UTC)
if mtime > since:
changed_files.append(item)
except (OSError, ValueError):
# Skip files we can't stat
continue
return changed_files
def create_archive(self, timestamp: Optional[str] = None) -> Path | None:
if not self.config.backup_source_dir.exists():
raise FileNotFoundError(f"backup source directory does not exist: {self.config.backup_source_dir}")
stamp = timestamp or utc_now().strftime("%Y%m%d-%H%M%S")
self.config.work_dir.mkdir(parents=True, exist_ok=True)
# Clean up stale checkpoints from previous interrupted uploads before starting
self._cleanup_stale_checkpoints()
# Get remote backup chain info FIRST (source of truth)
# This ensures continuity even if local metadata is lost
remote_chain_info = self._get_remote_backup_chain_info()
# CRITICAL: Call _should_do_full_backup FIRST with caching
# This must happen BEFORE _apply_dynamic_strategy to ensure consistent decisions
do_full_backup = self._should_do_full_backup(use_cached=False)
print(f"[BACKUP] Full backup decision: {do_full_backup} (cached=False call)")
# Apply dynamic backup strategy (pass already-computed decision)
strategy = self._apply_dynamic_strategy(force_full_backup=do_full_backup)
# Check if backup should be skipped
if strategy.get("skip_backup", False):
print(f"[Dynamic Strategy] Backup skipped: {strategy.get('reason', 'Unknown')}")
return None
# Apply dynamic settings
dynamic_compression = strategy.get("compression_level", self.config.compression_level)
dynamic_split = strategy.get("split_size", self.config.split_size)
# Temporarily override config for this backup
original_compression = self.config.compression_level
original_split = self.config.split_size
self.config.compression_level = dynamic_compression
self.config.split_size = dynamic_split
try:
# Use cached result from earlier call
last_backup_time = self._get_last_backup_time()
# Get parent_meta_path and chain_id from remote chain info
parent_meta_path = None
chain_id = None
if remote_chain_info:
chain_id = remote_chain_info.get("chain_id")
if do_full_backup:
parent_meta_path = None # Full backup starts new chain
print(f"[BACKUP] Creating new chain (full backup), chain_id={chain_id}")
else:
# For incremental backup: find the latest backup in the chain
# Use chain_id to find all backups, the latest one is the parent
chain = self._find_all_backups_by_chain_id(chain_id, repo_id=self.config.dataset_repo)
if chain and len(chain) > 0:
parent_meta_path = chain[-1] # Last one is the latest
print(f"[BACKUP] Remote chain: parent={parent_meta_path}, chain_id={chain_id}")
else:
# Fallback: use remote_chain_info's parent_meta_path
# This handles cases where we can't enumerate all backups
parent_meta_path = remote_chain_info.get("parent_meta_path")
if parent_meta_path:
print(f"[BACKUP] Remote chain (fallback): parent={parent_meta_path}, chain_id={chain_id}")
else:
print(f"[BACKUP] Warning: No parent found, starting new chain, chain_id={chain_id}")
archive = self.config.work_dir / f"openclaw-backup-{stamp}.tar.gz"
file_count = 0
# Use configured compression level (1-9, where 9 is maximum compression)
# For large files (700MB+), level 6 provides good balance
compression_level = max(1, min(9, self.config.compression_level))
with tarfile.open(archive, "w:gz", compresslevel=compression_level) as tar:
if do_full_backup or last_backup_time is None:
print(f"[BACKUP] Creating FULL backup: {archive.name}")
print(f"[BACKUP] ============================================")
print(f"[BACKUP] Backup source: {self.config.backup_source_dir}")
# Backup main state directory
if self.config.backup_source_dir.exists():
state_files = [f for f in self.config.backup_source_dir.rglob("*") if f.is_file()]
tar.add(str(self.config.backup_source_dir), arcname="openclaw-state")
file_count += len(state_files)
print(f"[BACKUP] ✓ openclaw-state: {len(state_files)} files")
else:
print(f"[BACKUP] ✗ openclaw-state: directory not found!")
# Backup extra directories
for archive_root, local_dir in self._extra_root_dirs().items():
if local_dir.exists():
extra_files = [f for f in local_dir.rglob("*") if f.is_file()]
tar.add(str(local_dir), arcname=archive_root)
file_count += len(extra_files)
print(f"[BACKUP] ✓ {archive_root}: {len(extra_files)} files")
else:
print(f"[BACKUP] - {archive_root}: directory not found (skipped)")
else:
print(f"[BACKUP] Creating INCREMENTAL backup: {archive.name}")
print(f"[BACKUP] ============================================")
print(f"[BACKUP] Last backup: {last_backup_time.isoformat()}")
# Add changed files from main state directory
changed_files = self._get_changed_files(self.config.backup_source_dir, last_backup_time)
for file_path in changed_files:
arcname = f"openclaw-state/{file_path.relative_to(self.config.backup_source_dir)}"
tar.add(str(file_path), arcname=arcname)
file_count += len(changed_files)
print(f"[BACKUP] ✓ openclaw-state: {len(changed_files)} changed files")
if changed_files:
for fp in changed_files[:5]:
print(f"[BACKUP] - {fp.name}")
if len(changed_files) > 5:
print(f"[BACKUP] ... and {len(changed_files) - 5} more")
# Add changed files from extra directories
for archive_root, local_dir in self._extra_root_dirs().items():
if local_dir.exists():
changed_files = self._get_changed_files(local_dir, last_backup_time)
for file_path in changed_files:
arcname = f"{archive_root}/{file_path.relative_to(local_dir)}"
tar.add(str(file_path), arcname=arcname)
file_count += len(changed_files)
print(f"[BACKUP] ✓ {archive_root}: {len(changed_files)} changed files")
if changed_files:
for fp in changed_files[:5]:
print(f"[BACKUP] - {fp.name}")
if len(changed_files) > 5:
print(f"[BACKUP] ... and {len(changed_files) - 5} more")
else:
print(f"[BACKUP] - {archive_root}: directory not found (skipped)")
# Always add extra files (they're typically small config files)
for archive_name, local_file in self.config.extra_files.items():
if local_file.exists() and local_file.is_file():
tar.add(str(local_file), arcname=archive_name)
file_count += 1
# Save metadata for next backup
is_full_backup = do_full_backup or last_backup_time is None
# Determine parent_meta_path and chain_id from remote chain info (source of truth)
# For incremental backups, parent is the previous backup's .meta.json
# For full backups, parent is None and new chain_id is generated
if not is_full_backup and remote_chain_info:
# Use parent_meta_path and chain_id from remote - already set above
pass # parent_meta_path and chain_id already set from remote_chain_info
elif is_full_backup:
parent_meta_path = None # Full backup starts new chain
chain_id = None # Will be generated in _save_backup_metadata
self._save_backup_metadata(archive, file_count, is_full_backup, parent_meta_path, chain_id)
# Check archive size and warn if too large
archive_size_mb = archive.stat().st_size / (1024 * 1024)
if archive_size_mb > self.config.size_warning_mb:
print(f"⚠️ WARNING: Backup size ({archive_size_mb:.1f}MB) exceeds warning threshold ({self.config.size_warning_mb}MB)")
print(f" Consider: 1) Excluding large files 2) Using split archives 3) Reducing backup frequency")
if is_full_backup:
print(f"Full backup complete: {file_count} files, {archive_size_mb:.1f}MB")
else:
print(f"Incremental backup complete: {file_count} files, {archive_size_mb:.1f}MB")
# Encrypt archive if enabled (before split)
if self.config.encryption_enabled and self.config.encryption_password:
result = self._encrypt_archive(archive)
if isinstance(result, Path):
enc_checksum = self._calculate_checksum(result)
enc_size = result.stat().st_size
enc_volumes = [result.name]
self._update_volumes_in_metadata(enc_volumes, f"sha256:{enc_checksum}", enc_size)
else:
result = archive
# Split archive if configured and file is large
# Note: checksum must be calculated BEFORE split for non-encrypted backups
if self.config.split_size and archive_size_mb > 500:
# For non-encrypted backups, calculate checksum before split
if not (self.config.encryption_enabled and self.config.encryption_password):
pre_split_checksum = self._calculate_checksum(result)
pre_split_size = result.stat().st_size
result = self._split_archive(result)
# Update metadata with actual volumes (split files)
# This must happen AFTER split since _save_backup_metadata is called BEFORE split
if isinstance(result, list):
is_encrypted = self.config.encryption_enabled and self.config.encryption_password
if is_encrypted:
# For encrypted + split backups, checksum was already stored when encryption
# happened (before split). After split, we only update volumes list.
# Checksum verification will happen after merge+decrypt in restore flow.
self._update_volumes_in_metadata([v.name for v in result])
else:
# For non-encrypted split backups, use checksum calculated before split
total_size = sum(v.stat().st_size for v in result)
self._update_volumes_in_metadata(
[v.name for v in result],
checksum=f"sha256:{pre_split_checksum}",
archive_size=total_size
)
return result
finally:
# Restore original config
self.config.compression_level = original_compression
self.config.split_size = original_split
def _encrypt_archive(self, archive: Path) -> Path:
"""Encrypt archive using AES-256-CBC with password.
Args:
archive: Path to the archive to encrypt
Returns:
Path to the encrypted archive (.tar.gz.enc)
"""
if not self.config.encryption_password:
raise ValueError("Encryption password not set")
encrypted_path = archive.with_suffix(archive.suffix + ".enc")
print(f"Encrypting archive: {archive.name} -> {encrypted_path.name}")
try:
import subprocess
import tempfile
import os as _os
with tempfile.NamedTemporaryFile(mode="w", delete=False) as pw_file:
_os.chmod(pw_file.file.fileno(), 0o600)
pw_file.write(self.config.encryption_password)
pw_file.flush()
pw_file_path = pw_file.name
try:
result = subprocess.run(
[
"openssl", "enc", "-aes-256-cbc", "-salt", "-pbkdf2",
"-in", str(archive),
"-out", str(encrypted_path),
"-pass", f"file:{pw_file_path}",
],
capture_output=True,
text=True,
check=True,
)
finally:
_os.unlink(pw_file_path)
archive.unlink()
print(f"Encrypted: {encrypted_path.name} ({encrypted_path.stat().st_size / (1024*1024):.1f}MB)")
return encrypted_path
except subprocess.CalledProcessError as e:
raise RuntimeError(f"Encryption failed: {e.stderr}")
def _decrypt_archive(self, encrypted_path: Path, target_path: Path) -> Path:
"""Decrypt archive using AES-256-CBC with password.
Args:
encrypted_path: Path to the encrypted archive (.tar.gz.enc)
target_path: Path where decrypted archive should be written
Returns:
Path to the decrypted archive
"""
if not self.config.encryption_password:
raise ValueError("Decryption password not set")
print(f"Decrypting archive: {encrypted_path.name} -> {target_path.name}")
try:
import subprocess
import tempfile
import os as _os
with tempfile.NamedTemporaryFile(mode="w", delete=False) as pw_file:
_os.chmod(pw_file.file.fileno(), 0o600)
pw_file.write(self.config.encryption_password)
pw_file.flush()
pw_file_path = pw_file.name
try:
result = subprocess.run(
[
"openssl", "enc", "-d", "-aes-256-cbc", "-pbkdf2",
"-in", str(encrypted_path),
"-out", str(target_path),
"-pass", f"file:{pw_file_path}",
],
capture_output=True,
text=True,
check=True,
)
finally:
_os.unlink(pw_file_path)
print(f"Decrypted: {target_path.name} ({target_path.stat().st_size / (1024*1024):.1f}MB)")
return target_path
except subprocess.CalledProcessError as e:
raise RuntimeError(f"Decryption failed: {e.stderr}")
def _split_archive(self, archive: Path) -> list[Path]:
"""Split large archive into smaller volumes."""
import subprocess
print(f"Splitting archive into {self.config.split_size} volumes...")
# Parse split size (e.g., "500M" -> 500)
split_mb = 500
if self.config.split_size:
size_str = self.config.split_size.upper()
if size_str.endswith('M'):
split_mb = int(size_str[:-1])
elif size_str.endswith('G'):
split_mb = int(size_str[:-1]) * 1024
split_bytes = split_mb * 1024 * 1024
# Use split command to create volumes
base_name = archive.name
volumes_pattern = self.config.work_dir / f"{base_name}.part-"
try:
result = subprocess.run(
["split", "-b", str(split_bytes), str(archive), str(volumes_pattern)],
capture_output=True,
text=True,
check=True
)
# Remove original archive after splitting
archive.unlink()
# Return list of volume files
volumes = sorted(self.config.work_dir.glob(f"{base_name}.part-*"))
print(f"Created {len(volumes)} volume(s)")
for vol in volumes:
vol_size_mb = vol.stat().st_size / (1024 * 1024)
print(f" - {vol.name}: {vol_size_mb:.1f}MB")
return volumes
except subprocess.CalledProcessError as e:
print(f"Warning: Failed to split archive: {e}")
print(f"Continuing with unsplit archive")
return [archive]
except FileNotFoundError:
print(f"Warning: 'split' command not found. Install coreutils.")
print(f"Continuing with unsplit archive")
return [archive]
def _upload_file_with_retry(self, path_or_fileobj, path_in_repo: str, commit_message: str, max_retries: int = 3, repo_id: Optional[str] = None) -> None:
"""Upload file with retry logic for rate limiting.
Args:
path_or_fileobj: File path or file object to upload
path_in_repo: Path in the repository
commit_message: Commit message
max_retries: Maximum number of retries
repo_id: Optional repository ID to upload to. If not specified, uses dataset_repo
"""
upload_repo = repo_id or self.config.dataset_repo
for attempt in range(max_retries):
try:
self.api.upload_file(
repo_id=upload_repo,
repo_type=self.config.repo_type,
path_or_fileobj=path_or_fileobj,
path_in_repo=path_in_repo,
commit_message=commit_message,
)
return
except Exception as e:
error_str = str(e)
# Check for rate limit error (429)
if "429" in error_str or "Too Many Requests" in error_str:
if attempt < max_retries - 1:
# Extract retry wait time from error message
match = re.search(r'Retry after (\d+) seconds', error_str)
if match:
wait_time = int(match.group(1))
print(f"Rate limit hit (429), waiting {wait_time} seconds (server requested) before retry {attempt + 1}/{max_retries}...")
else:
# Exponential backoff: 60, 120, 240 seconds
wait_time = 60 * (2 ** attempt)
print(f"Rate limit hit (429), waiting {wait_time} seconds (exponential backoff) before retry {attempt + 1}/{max_retries}...")
time.sleep(wait_time)
continue
# Re-raise if not rate limit or max retries reached
raise
def upload_backup(self, archive: Path | list[Path]) -> str:
if isinstance(archive, list):
volumes = archive
is_split = True
else:
volumes = [archive]
is_split = False
# Get current backup metadata - first from local, then from remote if not found
backup_meta = self._get_backup_metadata()
if not backup_meta:
# Try to get metadata from remote (in case local cache was lost)
remote_chain_info = self._get_remote_backup_chain_info()
if remote_chain_info:
# Build minimal metadata from remote chain info
backup_meta = {
"chain_id": remote_chain_info.get("chain_id"),
"parent": remote_chain_info.get("parent_meta_path"),
"backup_type": remote_chain_info.get("backup_type"),
}
print(f"[UPLOAD] Using remote chain info: chain_id={backup_meta.get('chain_id')}")
else:
raise RuntimeError("Backup metadata not found. Cannot upload.")
# Generate meta.json name early (needed for checkpoint)
first_volume_name = volumes[0].name
base_name = first_volume_name
if ".part-" in base_name:
base_name = base_name.rsplit(".part-", 1)[0]
meta_json_name = f"{base_name}.meta.json"
if self.config.path_prefix:
meta_json_remote = f"{self.config.path_prefix}/{meta_json_name}"
else:
meta_json_remote = meta_json_name
# Create checkpoint file for recovery - upload BEFORE any volume to establish a recovery point
checkpoint_json = {
"status": "uploading",
"backup_meta_path": meta_json_remote,
"volumes": [vol.name for vol in volumes],
"volumes_uploaded": [],
"created_at_utc": utc_now().isoformat(timespec="seconds"),
}
checkpoint_name = f"{base_name}.checkpoint.json"
if self.config.path_prefix:
checkpoint_remote = f"{self.config.path_prefix}/{checkpoint_name}"
else:
checkpoint_remote = checkpoint_name
print(f"Creating checkpoint file: {checkpoint_name}")
self._upload_file_with_retry(
path_or_fileobj=(json.dumps(checkpoint_json, ensure_ascii=False, indent=2) + "\n").encode("utf-8"),
path_in_repo=checkpoint_remote,
commit_message=f"backup: {checkpoint_name} (checkpoint)",
)
# Upload all volumes, updating checkpoint after each
remote_volumes = []
for vol in volumes:
if self.config.path_prefix:
remote_path = f"{self.config.path_prefix}/{vol.name}"
else:
remote_path = vol.name
print(f"Uploading {vol.name}...")
self._upload_file_with_retry(
path_or_fileobj=str(vol),
path_in_repo=remote_path,
commit_message=f"backup: {vol.name}",
)
remote_volumes.append(remote_path)
# Update checkpoint after each volume upload (for recovery)
checkpoint_json["volumes_uploaded"].append(vol.name)
print(f"Updating checkpoint: {vol.name} uploaded")
self._upload_file_with_retry(
path_or_fileobj=(json.dumps(checkpoint_json, ensure_ascii=False, indent=2) + "\n").encode("utf-8"),
path_in_repo=checkpoint_remote,
commit_message=f"backup: {checkpoint_name} (checkpoint update)",
)
# Update backup_meta with remote paths for volumes
# Note: volumes in .meta.json should be local names (without path_prefix)
# because restore code will prepend path_prefix when downloading
backup_meta["volumes"] = [vol.name for vol in volumes] # Local names only
backup_meta["is_split"] = is_split
# Generate and upload per-backup .meta.json
print(f"Uploading {meta_json_name}...")
self._upload_file_with_retry(
path_or_fileobj=(json.dumps(backup_meta, ensure_ascii=False, indent=2) + "\n").encode("utf-8"),
path_in_repo=meta_json_remote,
commit_message=f"backup: {meta_json_name}",
)
# Update latest-backup.json to point to the new .meta.json
latest_backup_json = {
"dataset": self.config.dataset_repo,
"latest": meta_json_remote, # Full path with path_prefix
"created_at_utc": utc_now().isoformat(timespec="seconds"),
}
# latest-backup.json should always be uploaded to dataset root (same level as backups/)
latest_json_path = "latest-backup.json"
print("Updating latest-backup.json...")
self._upload_file_with_retry(
path_or_fileobj=(json.dumps(latest_backup_json, ensure_ascii=False, indent=2) + "\n").encode("utf-8"),
path_in_repo=latest_json_path,
commit_message="backup: update latest-backup.json",
)
# Delete checkpoint file now that backup is complete
print(f"Deleting checkpoint file: {checkpoint_name}")
self._delete_checkpoint_with_retry(checkpoint_remote)
def _delete_checkpoint_with_retry(self, checkpoint_remote: str, max_retries: int = 3) -> None:
"""Delete checkpoint file with retry logic.
Args:
checkpoint_remote: Path to checkpoint file in repo
max_retries: Maximum deletion attempts
"""
for attempt in range(max_retries):
try:
self.api.delete_file(
path_in_repo=checkpoint_remote,
repo_id=self.config.dataset_repo,
repo_type=self.config.repo_type,
)
print(f"Checkpoint file deleted successfully")
return
except Exception as e:
error_str = str(e)
if "404" in error_str or "Entry Not Found" in error_str:
print(f"Checkpoint file already deleted or never existed")
return
if attempt < max_retries - 1:
wait_time = 2 ** attempt
print(f"Warning: Could not delete checkpoint file (attempt {attempt + 1}/{max_retries}): {e}")
print(f"Retrying in {wait_time}s...")
time.sleep(wait_time)
else:
print(f"Warning: Could not delete checkpoint file after {max_retries} attempts: {e}")
print(f"The checkpoint will be cleaned up during the next prune operation")
raise
def _cleanup_stale_checkpoints(self) -> None:
"""Clean up stale checkpoint files from previous interrupted uploads.
A checkpoint is considered stale if:
- It is older than 24 hours
- It has no corresponding backup archive uploaded
"""
try:
files = self.api.list_repo_files(
repo_id=self.config.dataset_repo,
repo_type=self.config.repo_type,
)
cutoff_time = utc_now() - dt.timedelta(hours=24)
cleaned = 0
for path in files:
if not isinstance(path, str):
continue
if not path.endswith(".checkpoint.json"):
continue
try:
with tempfile.TemporaryDirectory() as tmp_dir:
checkpoint_local = self._download_backup(path, Path(tmp_dir))
with open(checkpoint_local, "r") as f:
checkpoint_data = json.load(f)
created_at = checkpoint_data.get("created_at_utc", "")
if created_at:
checkpoint_time = dt.datetime.fromisoformat(created_at)
if checkpoint_time < cutoff_time:
self.api.delete_file(
path_in_repo=path,
repo_id=self.config.dataset_repo,
repo_type=self.config.repo_type,
)
print(f"[CLEANUP] Deleted stale checkpoint: {path}")
cleaned += 1
except Exception:
match = re.search(r"(\d{8}-\d{6})", path)
if match:
try:
file_time = dt.datetime.strptime(match.group(1), "%Y%m%d-%H%M%S")
if file_time < cutoff_time:
self.api.delete_file(
path_in_repo=path,
repo_id=self.config.dataset_repo,
repo_type=self.config.repo_type,
)
print(f"[CLEANUP] Deleted stale checkpoint (by filename): {path}")
cleaned += 1
except (ValueError, Exception):
pass
if cleaned > 0:
print(f"[CLEANUP] Cleaned up {cleaned} stale checkpoint(s)")
except Exception as e:
print(f"[CLEANUP] Warning: Could not clean up stale checkpoints: {e}")
def _verify_upload(self, files_to_verify: list[str], repo_id: Optional[str] = None) -> None:
"""Verify that the uploaded files exist in the repository.
Uses lightweight API checks for large files and selective download
verification for small metadata files only.
Args:
files_to_verify: List of file paths to verify
repo_id: Optional repository ID to verify in. If not specified, uses dataset_repo
"""
files_to_verify = [f for f in files_to_verify if f]
if not files_to_verify:
raise RuntimeError("Upload verification failed: no files to verify")
verify_repo = repo_id or self.config.dataset_repo
time.sleep(2)
try:
repo_files = self.api.list_repo_files(
repo_id=verify_repo,
repo_type=self.config.repo_type,
)
repo_files_set = set(repo_files)
for path in files_to_verify:
if path not in repo_files_set:
raise RuntimeError(f"Upload verification failed: {path} not found in repository")
# For small metadata files (< 1MB), do actual download verification
# For large archive files, presence in list_repo_files is sufficient
if path.endswith(".meta.json") or path.endswith("latest-backup.json"):
try:
with tempfile.TemporaryDirectory() as tmp_dir:
self.api.hf_hub_download(
repo_id=verify_repo,
filename=path,
local_dir=tmp_dir,
repo_type=self.config.repo_type,
token=self.token,
)
except Exception:
raise RuntimeError(f"Upload verification failed: {path} exists but cannot be downloaded")
print(f" Verified: {path}")
except Exception as e:
raise RuntimeError(f"Upload verification failed: {e}")
def _list_backup_archive_paths(self) -> list[str]:
if not hasattr(self.api, "list_repo_files"):
return []
files = self.api.list_repo_files(
repo_id=self.config.dataset_repo,
repo_type=self.config.repo_type,
)
path_prefix = self.config.path_prefix.strip("/")
path_prefix_slash = f"{path_prefix}/" if path_prefix else ""
# Group by backup timestamp, collect all paths for each backup
backups_by_time: dict[str, list[str]] = {}
for path in files:
if not isinstance(path, str):
continue
relative_path = path
if path_prefix_slash:
if not path.startswith(path_prefix_slash):
continue
relative_path = path[len(path_prefix_slash) :]
if "/" in relative_path:
continue
match = self._ARCHIVE_NAME_RE.fullmatch(relative_path)
if match is None:
continue
timestamp = match.group(1)
if timestamp not in backups_by_time:
backups_by_time[timestamp] = []
backups_by_time[timestamp].append(path)
# Sort backups by timestamp descending
sorted_timestamps = sorted(backups_by_time.keys(), reverse=True)
# Expand back to full list, keeping all volumes for each backup
result = []
for timestamp in sorted_timestamps:
result.extend(backups_by_time[timestamp])
return result
def _find_meta_file_for_timestamp(self, timestamp: str, files: list, path_prefix_slash: str) -> str | None:
"""Find the meta.json file path for a backup timestamp.
Args:
timestamp: The backup timestamp (e.g., "20260425-013501")
files: List of files in the repository
path_prefix_slash: Path prefix with trailing slash
Returns:
The meta file path if found, None otherwise
"""
for path in files:
if not isinstance(path, str):
continue
if timestamp in path and path.endswith(".meta.json"):
return path
return None
def prune_old_backups(self) -> list[str]:
print("[PRUNE] Starting old backup cleanup...")
print(f"[PRUNE] Retention policy: keep_count={self.config.keep_count}, min_full_backups=12")
if not hasattr(self.api, "delete_file"):
print("[PRUNE] API does not support delete_file, skipping cleanup")
return []
keep_count = self.config.keep_count if self.config.keep_count >= 1 else 48
min_full_backups = 12
if not hasattr(self.api, "list_repo_files"):
print("[PRUNE] API does not support list_repo_files, skipping cleanup")
return []
print("[PRUNE] Fetching file list from remote repository...")
files = self.api.list_repo_files(
repo_id=self.config.dataset_repo,
repo_type=self.config.repo_type,
)
print(f"[PRUNE] Found {len(files)} total files in repository")
path_prefix = self.config.path_prefix.strip("/")
path_prefix_slash = f"{path_prefix}/" if path_prefix else ""
backups_by_time: dict[str, list[str]] = {}
skipped_non_backup = 0
for path in files:
if not isinstance(path, str):
continue
relative_path = path
if path_prefix_slash:
if not path.startswith(path_prefix_slash):
continue
relative_path = path[len(path_prefix_slash) :]
if "/" in relative_path:
continue
match = self._ARCHIVE_NAME_RE.fullmatch(relative_path)
if match is None:
skipped_non_backup += 1
continue
timestamp = match.group(1)
if timestamp not in backups_by_time:
backups_by_time[timestamp] = []
backups_by_time[timestamp].append(path)
print(f"[PRUNE] Scanned files: {len(files)} total, {skipped_non_backup} non-backup files skipped")
print(f"[PRUNE] Identified {len(backups_by_time)} backup timestamp(s)")
if not backups_by_time:
print("[PRUNE] No backups found, nothing to prune")
return []
for ts, paths in sorted(backups_by_time.items(), reverse=True):
print(f"[PRUNE] Backup {ts}: {len(paths)} volume(s)")
sorted_timestamps = sorted(backups_by_time.keys(), reverse=True)
full_backup_timestamps = set()
for ts in sorted_timestamps:
meta_file = self._find_meta_file_for_timestamp(ts, files, path_prefix_slash)
if meta_file:
try:
cached_meta = self._get_cached_meta(meta_file)
if cached_meta is not None:
meta = cached_meta
else:
local_meta = self._download_backup(meta_file, self.config.work_dir)
with open(local_meta, "r") as f:
meta = json.load(f)
self._set_cached_meta(meta_file, meta)
if meta.get("backup_type") == "full":
full_backup_timestamps.add(ts)
except Exception:
pass
print(f"[PRUNE] Identified {len(full_backup_timestamps)} full backup(s), ensuring minimum 12 are kept")
if len(full_backup_timestamps) < min_full_backups:
print(f"[PRUNE] Not enough full backups ({len(full_backup_timestamps)}) to meet minimum {min_full_backups}, skipping cleanup")
return []
kept_full_backups_list = sorted(full_backup_timestamps, reverse=True)[:min_full_backups]
kept_full_backups = set(kept_full_backups_list)
oldest_kept_full = kept_full_backups_list[-1] if kept_full_backups_list else None
print(f"[PRUNE] Will keep {len(kept_full_backups)} newest full backups: {kept_full_backups_list[:3]}...")
if oldest_kept_full:
print(f"[PRUNE] Oldest kept full backup: {oldest_kept_full}")
old_full_timestamps = [ts for ts in sorted_timestamps if ts in full_backup_timestamps and ts not in kept_full_backups]
incremental_timestamps = [ts for ts in sorted_timestamps if ts not in full_backup_timestamps]
timestamps_to_delete_candidates = list(reversed(old_full_timestamps))
for ts in reversed(incremental_timestamps):
if oldest_kept_full and ts >= oldest_kept_full:
continue
timestamps_to_delete_candidates.append(ts)
timestamps_to_delete = timestamps_to_delete_candidates[:keep_count]
if not timestamps_to_delete:
print("[PRUNE] No backups to delete after applying retention policy")
return []
print(f"[PRUNE] Will delete {len(timestamps_to_delete)} backup(s): {len([t for t in timestamps_to_delete if t in full_backup_timestamps])} full, {len([t for t in timestamps_to_delete if t not in full_backup_timestamps])} incremental")
print(f"[PRUNE] Oldest backup to delete: {timestamps_to_delete[-1]}")
# Limit batch size to avoid rate limiting (max 50 files per cleanup)
max_files_per_cleanup = 50
total_files_to_delete = 0
for ts in timestamps_to_delete:
total_files_to_delete += len(backups_by_time[ts])
# Add 1 for meta.json (approximate)
total_files_to_delete += 1
if total_files_to_delete > max_files_per_cleanup:
# Limit to oldest backups first, but ensure at least one backup is cleaned
limited_timestamps = []
limited_count = 0
for ts in reversed(timestamps_to_delete): # Oldest first
ts_files = len(backups_by_time[ts]) + 1
if limited_count + ts_files <= max_files_per_cleanup or not limited_timestamps:
limited_timestamps.append(ts)
limited_count += ts_files
else:
break
timestamps_to_delete = list(reversed(limited_timestamps)) # Back to oldest-last order
print(f"[PRUNE] Limiting cleanup to {len(timestamps_to_delete)} backup(s) ({limited_count} files) to avoid rate limiting")
to_delete = []
seen_paths = set()
for timestamp in timestamps_to_delete:
archive_paths = backups_by_time[timestamp]
for archive_path in archive_paths:
if archive_path not in seen_paths:
to_delete.append(archive_path)
seen_paths.add(archive_path)
print(f"[PRUNE] Marking for deletion: backup {timestamp} ({len(archive_paths)} volume(s))")
meta_added = False
for archive_path in archive_paths:
archive_name = Path(archive_path).name
if ".part-" in archive_name:
base_name = archive_name.rsplit(".part-", 1)[0]
else:
base_name = archive_name
if base_name.endswith(".tar.gz"):
meta_name = base_name[:-7] + ".meta.json"
else:
meta_name = base_name + ".meta.json"
if path_prefix_slash:
meta_path = f"{path_prefix_slash}{meta_name}"
else:
meta_path = meta_name
if meta_path in files and meta_path not in seen_paths:
to_delete.append(meta_path)
seen_paths.add(meta_path)
if not meta_added:
print(f"[PRUNE] + meta.json: {meta_name}")
meta_added = True
# Also add checkpoint file if exists
checkpoint_name = base_name + ".checkpoint.json"
if path_prefix_slash:
checkpoint_path = f"{path_prefix_slash}{checkpoint_name}"
else:
checkpoint_path = checkpoint_name
if checkpoint_path in files and checkpoint_path not in seen_paths:
to_delete.append(checkpoint_path)
seen_paths.add(checkpoint_path)
print(f"[PRUNE] + checkpoint.json: {checkpoint_name}")
# Find orphaned checkpoint files (checkpoint without corresponding backup)
orphaned_checkpoints = []
for path in files:
if not isinstance(path, str):
continue
if not path.endswith(".checkpoint.json"):
continue
# Extract base name from checkpoint
checkpoint_name = Path(path).name
base_name = checkpoint_name[:-len(".checkpoint.json")]
# Check if corresponding backup exists
has_backup = False
for ts, archive_paths in backups_by_time.items():
for ap in archive_paths:
ap_name = Path(ap).name
if ".part-" in ap_name:
ap_base = ap_name.rsplit(".part-", 1)[0]
else:
ap_base = ap_name
if ap_base == base_name:
has_backup = True
break
if has_backup:
break
if not has_backup and path not in seen_paths:
orphaned_checkpoints.append(path)
seen_paths.add(path)
if orphaned_checkpoints:
print(f"[PRUNE] Found {len(orphaned_checkpoints)} orphaned checkpoint(s)")
for cp in orphaned_checkpoints:
to_delete.append(cp)
print(f"[PRUNE] + orphaned checkpoint: {Path(cp).name}")
print(f"[PRUNE] Total files to delete: {len(to_delete)}")
prune_repo = self.config.dataset_repo
deleted_count = 0
skipped_count = 0
rate_limit_hits = 0
import time
for path in to_delete:
print(f"[PRUNE] Deleting ({deleted_count + skipped_count + 1}/{len(to_delete)}): {path}")
retry_count = 0
max_retries = 3
while retry_count < max_retries:
try:
self.api.delete_file(
repo_id=prune_repo,
repo_type=self.config.repo_type,
path_in_repo=path,
commit_message=f"backup: prune {Path(path).name}",
)
deleted_count += 1
break
except Exception as e:
error_str = str(e)
if "404" in error_str or "Entry Not Found" in error_str:
print(f"[PRUNE] ⚠ File already gone or never existed, skipping: {path}")
skipped_count += 1
break
elif "429" in error_str or "Too Many Requests" in error_str:
retry_count += 1
rate_limit_hits += 1
if retry_count >= max_retries:
print(f"[PRUNE] ✗ Rate limit exceeded after {max_retries} retries, stopping cleanup: {path}")
print(f"[PRUNE] Cleanup interrupted due to rate limiting. Deleted {deleted_count} file(s), skipped {skipped_count} (not found), rate limit hits: {rate_limit_hits}")
return to_delete[:deleted_count + skipped_count]
wait_time = min(2 ** retry_count * 5, 60)
print(f"[PRUNE] ⚠ Rate limit hit ({retry_count}/{max_retries}), waiting {wait_time}s before retry...")
time.sleep(wait_time)
continue
else:
print(f"[PRUNE] ✗ Delete failed: {e}")
raise
print(f"[PRUNE] Cleanup complete, deleted {deleted_count} file(s), skipped {skipped_count} (not found), rate limit hits: {rate_limit_hits}")
return to_delete
def _manage_local_backups(self, archive: Path | list[Path]) -> None:
"""Manage local backup files based on retention policy."""
if not self.config.keep_local_backup:
# Clean up temporary files in work_dir
if isinstance(archive, list):
for vol in archive:
if vol.exists():
vol.unlink()
print(f"Cleaned up temporary volume: {vol.name}")
else:
if archive.exists():
archive.unlink()
print(f"Cleaned up temporary archive: {archive.name}")
return
# Keep local backups - move to local_backup_dir if specified
target_dir = self.config.local_backup_dir or self.config.work_dir
target_dir.mkdir(parents=True, exist_ok=True)
# Move/copy archives to target directory
archives_to_keep = []
if isinstance(archive, list):
for vol in archive:
target_path = target_dir / vol.name
if target_dir != self.config.work_dir:
shutil.copy2(vol, target_path)
vol.unlink()
archives_to_keep.append(target_path)
else:
target_path = target_dir / archive.name
if target_dir != self.config.work_dir:
shutil.copy2(archive, target_path)
archive.unlink()
archives_to_keep.append(target_path)
# Prune old local backups
self._prune_local_backups(target_dir)
print(f"Local backup kept: {target_dir}")
def _prune_local_backups(self, directory: Path) -> None:
"""Remove old local backups keeping only keep_local_count most recent."""
# Find all backup archives in directory
archives = sorted(
directory.glob("openclaw-backup-*.tar.gz*"),
key=lambda p: p.stat().st_mtime,
reverse=True
)
# Group by base name (for split archives)
base_names = {}
for arch in archives:
# Extract base name (remove .part-xx if present)
base_name = arch.name
if ".part-" in base_name:
base_name = base_name.rsplit(".part-", 1)[0]
if base_name not in base_names:
base_names[base_name] = []
base_names[base_name].append(arch)
# Keep only keep_local_count most recent backup sets
keep_count = self.config.keep_local_count
sorted_bases = sorted(base_names.keys(), reverse=True)
for base_name in sorted_bases[keep_count:]:
for arch in base_names[base_name]:
arch.unlink()
print(f"Pruned old local backup: {arch.name}")
def _cleanup_stale_lock(self, lock_path: Path) -> bool:
"""Clean up stale lock file from previous crashed process.
Returns True if lock was cleaned up or no lock existed.
Returns False if lock exists and is held by a live process.
"""
if not lock_path.exists():
return True
try:
with open(lock_path, "r") as f:
lines = f.readlines()
if not lines:
lock_path.unlink(missing_ok=True)
return True
try:
pid = int(lines[0].strip())
except (ValueError, IndexError):
lock_path.unlink(missing_ok=True)
return True
try:
os.kill(pid, 0)
return False
except ProcessLookupError:
print(f"[LOCK] Stale lock from dead process {pid}, cleaning up")
lock_path.unlink(missing_ok=True)
return True
except OSError:
return False
except Exception:
try:
lock_path.unlink(missing_ok=True)
except OSError:
pass
return True
def _acquire_lock(self, lock_path: Path) -> Optional[int]:
"""Acquire exclusive backup lock with stale lock cleanup and timeout."""
cleanup_result = self._cleanup_stale_lock(lock_path)
if not cleanup_result:
print("[LOCK] Lock file exists and is held by a live process, waiting...")
lock_fd = os.open(str(lock_path), os.O_CREAT | os.O_RDWR)
start_time = time.time()
timeout = 300
while True:
try:
fcntl.flock(lock_fd, fcntl.LOCK_EX | fcntl.LOCK_NB)
lock_uuid = str(uuid.uuid4())
os.write(lock_fd, f"{os.getpid()}\n{utc_now().isoformat()}\n{lock_uuid}\n".encode())
os.fsync(lock_fd)
self._lock_uuid = lock_uuid
return lock_fd
except Exception as e:
if time.time() - start_time > timeout:
os.close(lock_fd)
raise TimeoutError(f"Could not acquire backup lock after {timeout}s") from e
time.sleep(1)
def backup(self) -> Path | list[Path] | None:
lock_path = self.config.work_dir / "openclaw-backup.lock"
lock_path.parent.mkdir(parents=True, exist_ok=True)
lock_fd = None
result = False
try:
lock_fd = self._acquire_lock(lock_path)
return self._backup_inner()
except TimeoutError as e:
print(f"[LOCK] {e}, backup skipped")
return None
finally:
if lock_fd is not None:
try:
fcntl.flock(lock_fd, fcntl.LOCK_UN)
except (OSError, IOError):
pass
os.close(lock_fd)
try:
lock_path.unlink(missing_ok=True)
except OSError:
pass
def _backup_inner(self) -> Path | list[Path] | None:
self.ensure_dataset_repo()
if not self._verify_existing_backup():
print("backup aborted: cannot verify existing backup is restorable", file=sys.stderr)
print(" Please fix the restore issue before creating new backups.", file=sys.stderr)
return None
archive = self.create_archive()
if archive is None:
print("backup skipped by dynamic strategy")
return None
remote_path = self.upload_backup(archive)
print(f"backup uploaded: {self.config.dataset_repo}/{remote_path}")
self._manage_local_backups(archive)
return archive
def _verify_existing_backup(self) -> bool:
"""Verify that the existing latest backup metadata is valid.
Note: This method only validates metadata, not the actual backup files.
Full integrity verification is deferred to the restore process.
Returns True if:
- No existing backup (first backup)
- Existing backup metadata is valid
Returns False if:
- Existing backup metadata is invalid or corrupted
"""
print("Checking if existing backup metadata is valid...")
print(f" Checking backup target: {self.config.dataset_repo}")
verify_repo = self.config.dataset_repo
# Try to get metadata
try:
# latest-backup.json is ALWAYS at the dataset root (no path_prefix), not in backups/
meta_filename = "latest-backup.json"
print(f" Downloading {meta_filename}...")
meta_path = self._download_backup(meta_filename, self.config.work_dir)
print(f" Downloaded to {meta_path}")
with open(meta_path, "r") as f:
latest_meta = json.load(f)
print(f" Found backup metadata: {latest_meta.get('latest', 'unknown')}")
# Now download the actual .meta.json file to get volumes and is_split
actual_meta_path = latest_meta.get("latest", "")
if not actual_meta_path:
print(" No 'latest' entry found in latest-backup.json")
return False
print(f" Downloading actual meta file: {actual_meta_path}...")
cached_meta = self._get_cached_meta(actual_meta_path)
if cached_meta is not None:
backup_meta = cached_meta
print(f" Using cached metadata for {actual_meta_path}")
else:
actual_meta_local = self._download_backup(actual_meta_path, self.config.work_dir, repo_id=verify_repo)
with open(actual_meta_local, "r") as f:
backup_meta = json.load(f)
self._set_cached_meta(actual_meta_path, backup_meta)
is_split = backup_meta.get("is_split", False)
volumes = backup_meta.get("volumes", [])
print(f" backup_meta keys: {list(backup_meta.keys())}")
print(f" is_split: {is_split}")
print(f" volumes: {volumes}")
# Validate that volumes are specified
if not volumes:
print(" Warning: No volumes found in backup metadata")
return False
# Verify that backup files exist in the repository using API (no download)
print(" Verifying backup files exist in repository...")
try:
repo_files = self.api.list_repo_files(
repo_id=verify_repo,
repo_type=self.config.repo_type,
)
repo_files_set = set(repo_files)
for vol_path in volumes:
# Add path_prefix if not already present
check_path = vol_path
if self.config.path_prefix and not vol_path.startswith(self.config.path_prefix):
check_path = f"{self.config.path_prefix}/{vol_path}"
if check_path not in repo_files_set:
print(f" ✗ Backup file not found: {check_path}")
return False
print(f" ✓ Backup file exists: {check_path}")
print(f" ✓ All {len(volumes)} backup file(s) verified to exist")
except Exception as e:
print(f" Warning: Could not verify backup files exist: {e}")
# Continue anyway, don't fail the backup due to API issues
print(" Note: Full integrity verification will be performed during restore")
return True
except Exception as e:
# No metadata means no existing backup (first backup)
print(f" No existing backup found (first backup): {type(e).__name__}: {e}")
return True
@staticmethod
def replace_dir_contents(source_dir: Path, target_dir: Path) -> None:
target_dir.mkdir(parents=True, exist_ok=True)
for existing in target_dir.iterdir():
try:
if existing.is_dir() and not existing.is_symlink():
shutil.rmtree(existing)
else:
existing.unlink()
except (OSError, PermissionError):
pass
for item in source_dir.iterdir():
destination = target_dir / item.name
if item.is_symlink():
try:
if destination.exists() or destination.is_symlink():
destination.unlink()
except OSError:
pass
try:
destination.symlink_to(os.readlink(item))
except OSError:
pass
elif item.is_dir():
if destination.exists():
shutil.rmtree(destination)
shutil.copytree(item, destination, symlinks=True)
else:
try:
shutil.copy2(item, destination)
except FileExistsError:
try:
destination.unlink()
shutil.copy2(item, destination)
except Exception:
pass
def _download_backup(self, remote_path: str, local_dir: Path, repo_id: Optional[str] = None) -> Path:
"""Download a specific backup from the dataset.
First attempts to use hf_hub_download (with cache bypass), then falls back
to direct HTTP download via urllib. Always ensures latest version is obtained.
Uses in-memory caching for .meta.json files to avoid repeated downloads during
the same session.
Args:
remote_path: Path to the file in the repository
local_dir: Local directory to download to
repo_id: Optional repository ID to download from. If not specified,
uses dataset_repo as default
"""
# Determine which repo to download from
# Default to dataset_repo (caller should explicitly pass repo_id for correct dataset)
download_repo = repo_id or self.config.dataset_repo
# Check in-memory cache first (for .meta.json files to avoid repeated downloads)
if remote_path in self._meta_file_cache:
cached_path = self._meta_file_cache[remote_path]
if cached_path.exists():
return cached_path
local_dir = Path(local_dir)
local_dir.mkdir(parents=True, exist_ok=True)
local_file_path = local_dir / remote_path
if local_file_path.exists():
local_file_path.unlink()
if hf_hub_download is not None:
try:
print(f"[_download_backup] Trying hf_hub_download for {remote_path} from {download_repo}...")
downloaded_path = hf_hub_download(
repo_id=download_repo,
repo_type=self.config.repo_type,
filename=remote_path,
local_dir=str(local_dir),
force_download=True,
token=self.token,
)
print(f"[_download_backup] hf_hub_download succeeded: {downloaded_path}")
result_path = Path(downloaded_path)
if remote_path.endswith(".meta.json"):
self._meta_file_cache[remote_path] = result_path
return result_path
except Exception as e:
print(f"[_download_backup] hf_hub_download failed: {type(e).__name__}: {e}")
print(f"[_download_backup] Falling back to direct HTTP download...")
print(f"[_download_backup] Using direct HTTP for {remote_path} from {download_repo}...")
endpoint = f"https://huggingface.co/datasets/{download_repo}/resolve/main/{remote_path}"
headers = {}
if self.token:
headers["Authorization"] = f"Bearer {self.token}"
req = urllib.request.Request(endpoint, headers=headers)
with urllib.request.urlopen(req, timeout=60) as response:
local_file_path.write_bytes(response.read())
print(f"[_download_backup] Direct HTTP download succeeded: {local_file_path}")
if remote_path.endswith(".meta.json"):
self._meta_file_cache[remote_path] = local_file_path
return local_file_path
def _clear_hf_cache(self, remote_path: str) -> None:
"""Clear Hugging Face Hub cache for a specific file to force re-download."""
try:
# Get the default cache directory
cache_dir = Path.home() / ".cache" / "huggingface" / "hub"
if not cache_dir.exists():
return
# Construct the repo cache directory name
repo_name = self.config.dataset_repo.replace("/", "--")
repo_cache_dir = cache_dir / f"datasets--{repo_name}"
if not repo_cache_dir.exists():
return
# Find and remove cached files matching the remote_path
# The cache structure is: blobs/, refs/, snapshots/
# We need to find files in blobs/ that match our filename
filename = Path(remote_path).name
removed = False
# Search for cached files
for blob_file in repo_cache_dir.rglob("*"):
if blob_file.is_file() and blob_file.name == filename:
blob_file.unlink()
print(f" Cleared cache: {blob_file}")
removed = True
if not removed:
print(f" No cache found for: {filename}")
except Exception as e:
print(f" Warning: Could not clear cache: {e}")
def _get_cached_meta(self, meta_path: str) -> Optional[dict]:
"""Get cached metadata JSON, or None if not cached."""
return self._meta_cache.get(meta_path)
def _set_cached_meta(self, meta_path: str, meta: dict) -> None:
"""Cache metadata JSON."""
self._meta_cache[meta_path] = meta
def should_restore(self) -> bool:
"""Determine if restore should be performed based on configuration.
Restore logic:
- Container restart/rebuild always triggers restore
- This is by design to ensure data consistency
Users can configure OPENCLAW_RESTORE_DATASET_REPO via HF Space variables
to control restore behavior (defaults to OPENCLAW_BACKUP_DATASET_REPO).
Returns:
True (restore is always performed on container start)
"""
print("Restore will be performed (container restart/rebuild detected)")
return True
def _format_size(self, size_bytes: int) -> str:
"""Format byte size to human readable string."""
for unit in ["B", "KB", "MB", "GB", "TB"]:
if size_bytes < 1024:
return f"{size_bytes:.1f} {unit}"
size_bytes /= 1024
return f"{size_bytes:.1f} PB"
def _merge_directory(self, source: Path, target: Path) -> None:
"""Merge source directory into target directory.
For incremental backups: newer files overwrite older ones.
"""
target.mkdir(parents=True, exist_ok=True)
for item in source.rglob("*"):
if item.is_file():
rel_path = item.relative_to(source)
dest = target / rel_path
dest.parent.mkdir(parents=True, exist_ok=True)
try:
shutil.copy2(item, dest)
except FileExistsError:
dest.unlink()
shutil.copy2(item, dest)
elif item.is_dir():
rel_path = item.relative_to(source)
dest = target / rel_path
dest.mkdir(parents=True, exist_ok=True)
def restore(self) -> bool:
"""Restore from backup, automatically handling incremental backup chains.
Flow according to documentation:
1. Download latest-backup.json to get latest .meta.json path
2. Parse backup chain: download all .meta.json files, trace back to full backup
3. Download archives in order (from full to latest) and merge
4. Restore files to target directory
"""
lock_path = self.config.work_dir / "openclaw-backup.lock"
lock_path.parent.mkdir(parents=True, exist_ok=True)
lock_fd = None
result = False
try:
lock_fd = self._acquire_lock(lock_path)
if lock_fd is None:
print("restore skipped: a backup or restore is already in progress")
return False
os.write(lock_fd, f"{os.getpid()}\n{utc_now().isoformat()}\nrestore\n".encode())
os.fsync(lock_fd)
result = self._restore_inner()
except TimeoutError as e:
print(f"[LOCK] {e}, restore skipped")
return False
finally:
if lock_fd is not None:
try:
fcntl.flock(lock_fd, fcntl.LOCK_UN)
except (OSError, IOError):
pass
os.close(lock_fd)
try:
lock_path.unlink(missing_ok=True)
except Exception:
pass
if result:
restore_completed_flag = Path("/tmp/openclaw-restore-completed")
restore_completed_flag.touch()
print(f"Created restore completion flag: {restore_completed_flag}", file=sys.stderr)
return result
def _restore_inner(self) -> bool:
print("Starting restore process...")
restore_repo = self._get_restore_repo()
print(f"Restore source: {restore_repo}")
print(f"Target: {self.config.backup_source_dir}")
# Step 1: Determine starting point
# Check if user specified a specific backup via environment variable
specified_archive = os.getenv("OPENCLAW_RESTORE_ARCHIVE")
if specified_archive:
# User specified a backup archive name, add .meta.json suffix
meta_path = f"{specified_archive}.meta.json"
if self.config.path_prefix:
meta_path = f"{self.config.path_prefix}/{meta_path}"
print(f"Restoring specified backup: {specified_archive}")
else:
# Step 1: Download latest-backup.json from dataset root (always at root, regardless of path_prefix)
latest_json_path = "latest-backup.json"
try:
latest_json_local = self._download_backup(latest_json_path, self.config.work_dir, repo_id=restore_repo)
with open(latest_json_local, "r") as f:
latest_json = json.load(f)
meta_path = latest_json.get("latest", "")
print(f"Latest backup: {meta_path}")
except Exception as e:
print(f"Warning: latest-backup.json not found in dataset '{restore_repo}': {e}", file=sys.stderr)
print(f"This appears to be a newly initialized dataset", file=sys.stderr)
restore_skipped_flag = Path("/tmp/openclaw-restore-skipped-no-backup")
restore_skipped_flag.touch()
restore_completed_flag = Path("/tmp/openclaw-restore-completed")
restore_completed_flag.touch()
print(f"Created restore skipped flag: {restore_skipped_flag}", file=sys.stderr)
return True
# Step 2: Parse backup chain (download all .meta.json, trace back to full backup)
print("\n=== Step 2: Parsing backup chain ===")
chain_meta_paths = self._parse_backup_chain(meta_path, repo_id=restore_repo)
if not chain_meta_paths:
print("Error: Could not determine backup chain", file=sys.stderr)
return False
if len(chain_meta_paths) == 1:
print(f"Single backup (full): {chain_meta_paths[0]}")
else:
print(f"Backup chain with {len(chain_meta_paths)} backup(s):")
for i, mp in enumerate(chain_meta_paths, 1):
backup_type = "full" if i == 1 else "incremental"
print(f" [{i}] {mp} ({backup_type})")
# Step 3: Download archives in order and merge
print("\n=== Step 3: Downloading and merging archives ===")
extra_dirs = self._restore_root_dirs()
extra_files = self.config.extra_files
if extra_dirs:
print("Extra directories to restore:")
for archive_root, local_dir in extra_dirs.items():
print(f" - {archive_root} -> {local_dir}")
if extra_files:
print("Extra files to restore:")
for archive_name, local_file in extra_files.items():
print(f" - {archive_name} -> {local_file}")
restored_items = []
with tempfile.TemporaryDirectory(prefix="openclaw-restore-") as tmp:
tmp_dir = Path(tmp)
# Download and merge archives
merged_state = self._download_and_merge_chain(chain_meta_paths, tmp_dir, repo_id=restore_repo)
if not merged_state.exists() or not merged_state.is_dir():
print("Error: Merged state directory missing", file=sys.stderr)
return False
# Step 4: Restore files to target directory
print("\n=== Step 4: Restoring files ===")
# Restore main state
print(f"Restoring state to: {self.config.backup_source_dir}")
self.replace_dir_contents(merged_state / "openclaw-state", self.config.backup_source_dir)
restored_items.append(f"state -> {self.config.backup_source_dir}")
# Restore extra directories
for archive_root, local_dir in extra_dirs.items():
extracted_extra_dir = merged_state / archive_root
if extracted_extra_dir.exists():
if not extracted_extra_dir.is_dir():
print(f"Warning: backup archive {archive_root} is not a directory, skipping")
continue
print(f"Restoring {archive_root} to: {local_dir}")
self.replace_dir_contents(extracted_extra_dir, local_dir)
restored_items.append(f"{archive_root} -> {local_dir}")
# Restore extra files
for archive_name, local_file in extra_files.items():
extracted_file = merged_state / archive_name
if extracted_file.exists() and extracted_file.is_file():
local_file.parent.mkdir(parents=True, exist_ok=True)
shutil.copy2(str(extracted_file), str(local_file))
print(f"Restored extra file: {archive_name} -> {local_file}")
restored_items.append(f"{archive_name} -> {local_file}")
print("")
restore_repo = self._get_restore_repo()
print(f"✓ Restore completed successfully from {restore_repo}")
print(f" Chain: {len(chain_meta_paths)} backup(s) merged")
print(" Restored items:")
for item in restored_items:
print(f" - {item}")
restored_metadata = merged_state / "backup-info" / "last-backup-metadata.json"
if restored_metadata.exists():
self.config.metadata_dir.mkdir(parents=True, exist_ok=True)
target_metadata = self.config.metadata_dir / "last-backup-metadata.json"
shutil.copy2(str(restored_metadata), str(target_metadata))
print(f" Copied metadata to: {target_metadata}")
print("")
print("Backup restore completed successfully")
return True
def _parse_backup_chain(self, latest_meta_path: str, repo_id: Optional[str] = None) -> list[str]:
"""Parse backup chain by following parent pointers in .meta.json files.
Args:
latest_meta_path: Path to the latest .meta.json file (with path_prefix)
repo_id: Optional repository ID to download from. Uses dataset_repo if not specified.
Returns:
List of .meta.json paths in order from full backup to latest
"""
download_repo = repo_id or self.config.dataset_repo
chain = []
current_path = latest_meta_path
visited = set()
max_chain_length = 100
chain_id = None
latest_time = None
local_meta = self._download_backup(latest_meta_path, self.config.work_dir, repo_id=download_repo)
with open(local_meta, "r") as f:
latest_meta = json.load(f)
latest_time_str = latest_meta.get("created_at_utc") or latest_meta.get("last_backup_time")
if latest_time_str:
try:
latest_time = dt.datetime.fromisoformat(latest_time_str.replace("Z", "+00:00"))
except (ValueError, AttributeError):
latest_time = None
while current_path and len(chain) < max_chain_length:
if current_path in visited:
print(f"Warning: Circular reference detected in backup chain at {current_path}")
break
visited.add(current_path)
chain.append(current_path)
# Download and parse this .meta.json (use cache if available)
try:
cached_meta = self._get_cached_meta(current_path)
if cached_meta is not None:
meta = cached_meta
else:
local_meta = self._download_backup(current_path, self.config.work_dir, repo_id=download_repo)
with open(local_meta, "r") as f:
meta = json.load(f)
self._set_cached_meta(current_path, meta)
except Exception as e:
print(f"Error: Could not download/parse {current_path}: {e}", file=sys.stderr)
break
# Track chain_id from first meta
if chain_id is None:
chain_id = meta.get("chain_id")
# Get parent pointer
parent = meta.get("parent")
if not parent:
# Full backup, end of chain
if meta.get("backup_type") == "full":
# Normal full backup, end of chain
break
else:
# Incremental backup with null parent - chain is broken (parent field lost)
# Need to find the full backup by chain_id
print(f"Warning: Incremental backup has null parent (chain may be broken). chain_id={chain_id}")
break
# Build full path for parent (may need path_prefix)
if self.config.path_prefix and not parent.startswith(self.config.path_prefix):
parent = f"{self.config.path_prefix}/{parent}"
current_path = parent
if len(chain) >= max_chain_length:
print(f"Warning: Chain length exceeded safety limit ({max_chain_length})")
# If chain is broken (only incremental with null parent), try to find full backup
if len(chain) == 1 and chain_id:
first_meta_path = chain[0]
try:
local_meta = self._download_backup(first_meta_path, self.config.work_dir, repo_id=download_repo)
with open(local_meta, "r") as f:
first_meta = json.load(f)
if first_meta.get("backup_type") == "incremental" and not first_meta.get("parent"):
print(f"Attempting to rebuild chain using chain_id={chain_id}...")
all_backups = self._find_all_backups_by_chain_id(chain_id, repo_id=download_repo, latest_time=latest_time)
if all_backups:
print(f"Found {len(all_backups)} backup(s) in chain")
# all_backups is already sorted: [full, inc-1, inc-2, ...]
chain = all_backups
# Already in correct order (full first), no reverse needed
else:
print(f"Warning: Could not find backups for chain_id={chain_id}")
print(f" This incremental backup may not restore correctly without its full backup")
except Exception as e:
print(f"Warning: Could not check first meta for chain rebuild: {e}")
else:
# Normal case: reverse the chain built by following parent pointers
chain.reverse()
return chain
def _find_all_backups_by_chain_id(self, chain_id: str, repo_id: Optional[str] = None, latest_time: Optional[dt.datetime] = None) -> Optional[list[str]]:
"""Find all backups (full + incremental) for a given chain_id and sort by time.
Args:
chain_id: The chain ID to search for
repo_id: Optional repository ID to search in. Uses dataset_repo if not specified.
latest_time: Optional datetime to filter backups - only backups at or before this time are included
Returns:
List of .meta.json paths sorted by created_at_utc (earliest first), or None if no full backup found
"""
download_repo = repo_id or self.config.dataset_repo
try:
from huggingface_hub import HfApi
token = self.token
api = HfApi(token=token)
files = api.list_repo_files(repo_id=download_repo, repo_type=self.config.repo_type)
meta_files = [f for f in files if f.endswith(".meta.json")]
backups_with_time = []
full_backup_path = None
for meta_file in meta_files:
try:
cached_meta = self._get_cached_meta(meta_file)
if cached_meta is not None:
meta = cached_meta
else:
local_meta = self._download_backup(meta_file, self.config.work_dir, repo_id=download_repo)
with open(local_meta, "r") as f:
meta = json.load(f)
self._set_cached_meta(meta_file, meta)
if meta.get("chain_id") == chain_id:
created_at = meta.get("created_at_utc") or meta.get("last_backup_time")
if created_at:
try:
if isinstance(created_at, str):
created_at = dt.datetime.fromisoformat(created_at.replace("Z", "+00:00"))
except (ValueError, AttributeError):
created_at = None
if latest_time and created_at and created_at > latest_time:
continue
if meta.get("backup_type") == "full":
full_backup_path = (meta_file, created_at, meta)
else:
backups_with_time.append((meta_file, created_at, meta))
except Exception:
continue
if not full_backup_path:
print(f"Warning: No full backup found for chain_id={chain_id}")
return None
backups_with_time.sort(key=lambda x: x[1] if x[1] else dt.datetime.min)
chain = [full_backup_path[0]]
chain.extend([b[0] for b in backups_with_time])
return chain
except Exception as e:
print(f"Warning: Could not search for backups: {e}")
return None
def _download_and_merge_chain(self, chain_meta_paths: list[str], tmp_dir: Path, repo_id: Optional[str] = None) -> Path:
"""Download archives in chain order and merge them.
Args:
chain_meta_paths: List of .meta.json paths (from full to latest)
tmp_dir: Temporary directory for downloads and extraction
repo_id: Optional repository ID to download from. Uses dataset_repo if not specified.
Returns:
Path to merged state directory
"""
# Default to dataset_repo if not specified
download_repo = repo_id or self.config.dataset_repo
merged_dir = tmp_dir / "merged-state"
merged_dir.mkdir(parents=True, exist_ok=True)
print(f"Merging {len(chain_meta_paths)} backup(s) in chain...")
total_size = 0
successful_backups = 0
for i, meta_path in enumerate(chain_meta_paths):
backup_type = "full" if i == 0 else "incremental"
print(f" [{i+1}/{len(chain_meta_paths)}] Processing: {meta_path} ({backup_type})")
# Download and parse .meta.json (use cache if available)
try:
cached_meta = self._get_cached_meta(meta_path)
if cached_meta is not None:
meta = cached_meta
else:
local_meta = self._download_backup(meta_path, tmp_dir, repo_id=download_repo)
with open(local_meta, "r") as f:
meta = json.load(f)
self._set_cached_meta(meta_path, meta)
except Exception as e:
print(f"Error: Could not download/parse {meta_path}: {e}", file=sys.stderr)
if i == 0:
raise RuntimeError(f"Failed to download metadata for full backup: {meta_path}")
print(f" Warning: Stopping backup chain at {meta_path}, using {successful_backups} successful backup(s)")
break
is_split = meta.get("is_split", False)
is_encrypted = meta.get("encrypted", False)
volumes = meta.get("volumes", [])
expected_checksum = meta.get("checksum", "")
# Download archive(s)
archive_path = None
try:
if is_split and volumes:
# Download all split volumes and merge
print(f" Downloading {len(volumes)} split volumes...")
local_volumes = []
for vol_path in volumes:
# Add path_prefix if not already present
vol_remote_path = vol_path
if self.config.path_prefix and not vol_path.startswith(self.config.path_prefix):
vol_remote_path = f"{self.config.path_prefix}/{vol_path}"
local_vol = self._download_backup(vol_remote_path, tmp_dir, repo_id=download_repo)
local_volumes.append(local_vol)
# Merge volumes
# For encrypted splits, merged file should have .enc extension for proper decryption
if is_encrypted:
merged_archive = tmp_dir / f"merged-{i}.tar.gz.enc"
else:
merged_archive = tmp_dir / f"merged-{i}.tar.gz"
with open(merged_archive, "wb") as outfile:
for vol in sorted(local_volumes):
with open(vol, "rb") as infile:
outfile.write(infile.read())
archive_path = merged_archive
print(f" Merged {len(local_volumes)} volumes into {merged_archive.name}")
else:
# Single file archive
if volumes:
# Add path_prefix if not already present
vol_remote_path = volumes[0]
if self.config.path_prefix and not volumes[0].startswith(self.config.path_prefix):
vol_remote_path = f"{self.config.path_prefix}/{volumes[0]}"
archive_path = self._download_backup(vol_remote_path, tmp_dir, repo_id=download_repo)
else:
raise RuntimeError(f"No volumes found in metadata for {meta_path}")
# Verify checksum for all archives BEFORE decryption
# For split backups, verify the merged archive checksum if available
if expected_checksum:
try:
actual_checksum = self._calculate_checksum(archive_path)
if actual_checksum != expected_checksum.replace("sha256:", ""):
raise RuntimeError(
f"Checksum mismatch for {archive_path.name}: "
f"expected {expected_checksum}, got sha256:{actual_checksum}"
)
print(f" ✓ Checksum verified: {archive_path.name}")
except Exception as e:
print(f" Error: Integrity verification failed for {meta_path}: {e}", file=sys.stderr)
if i == 0:
raise RuntimeError(f"Full backup integrity check failed: {meta_path}")
print(f" Warning: Stopping backup chain at {meta_path}, using {successful_backups} successful backup(s)")
break
elif is_split:
print(f" ⚠ Checksum not available for split backup, skipping verification (create new backups to enable)")
if is_encrypted:
if not self.config.encryption_password:
raise RuntimeError(
f"Backup {meta_path} is encrypted but OPENCLAW_BACKUP_ENCRYPTION_PASSWORD is not set. "
f"Cannot restore encrypted backup without password."
)
decrypted_archive = tmp_dir / f"decrypted-{i}.tar.gz"
archive_path = self._decrypt_archive(archive_path, decrypted_archive)
except Exception as e:
print(f" Error: Failed to download archive for {meta_path}: {e}", file=sys.stderr)
if i == 0:
raise RuntimeError(f"Failed to download full backup: {meta_path}")
print(f" Warning: Stopping backup chain at {meta_path}, using {successful_backups} successful backup(s)")
break
# Checksum verification is now done BEFORE decryption (above)
# This ensures all archive types (encrypted, split, both) are verified consistently
# For split backups without expected_checksum, a warning was already printed above
archive_size = archive_path.stat().st_size
total_size += archive_size
print(f" Archive: {archive_path.name}, size: {self._format_size(archive_size)}")
# Extract
extract_dir = tmp_dir / f"extract-{i}"
extract_dir.mkdir(parents=True, exist_ok=True)
try:
with tarfile.open(archive_path, "r:gz") as tar:
tar.extractall(extract_dir)
except tarfile.ReadError as e:
print(f" Error: Failed to extract archive {archive_path}: {e}", file=sys.stderr)
if i == 0:
raise RuntimeError(f"Failed to extract full backup: {meta_path}")
print(f" Warning: Stopping backup chain at {meta_path}, using {successful_backups} successful backup(s)")
break
# Count files
file_count = sum(1 for _ in extract_dir.rglob("*") if _.is_file())
print(f" Files: {file_count}")
# Merge into merged directory
state_dir = extract_dir / "openclaw-state"
if state_dir.exists():
self._merge_directory(state_dir, merged_dir / "openclaw-state")
# Merge extra directories
for archive_root, _ in self._restore_root_dirs().items():
extra_dir = extract_dir / archive_root
if extra_dir.exists():
self._merge_directory(extra_dir, merged_dir / archive_root)
successful_backups += 1
if successful_backups == 0:
raise RuntimeError("No backups could be restored successfully")
print(f" Total downloaded: {self._format_size(total_size)}")
print(f" Successfully processed {successful_backups}/{len(chain_meta_paths)} backup(s)")
return merged_dir
def build_parser() -> argparse.ArgumentParser:
parser = argparse.ArgumentParser(description="Backup/restore OpenClaw state with Hugging Face Dataset")
parser.add_argument("action", choices=["backup", "restore"], help="action to perform")
return parser
def main(argv: Optional[list[str]] = None) -> int:
args = build_parser().parse_args(argv)
try:
config = BackupConfig.from_env()
except ValueError as exc:
print(f"skip {args.action}: {exc}", file=sys.stderr)
return 0
try:
runner = OpenClawBackup(config=config)
if args.action == "backup":
runner.backup()
return 0
# Restore action - always perform restore on container start
if args.action == "restore":
# Perform restore (container restart/rebuild always triggers restore)
success = runner.restore()
return 0 if success else 1
return 0
except Exception as exc: # pragma: no cover
print(f"{args.action} failed: {exc}", file=sys.stderr)
return 1
if __name__ == "__main__":
raise SystemExit(main())
|