Re: [BUG] fatal hang untarring 90GB file, possibly writebackrelated.

From: James Bottomley
Date: Thu Apr 28 2011 - 12:12:04 EST


On Thu, 2011-04-28 at 16:33 +0200, Jan Kara wrote:
> James, Colin, can you capture output of 'vmstat 1' while you do the
> copying? Thanks.

Sure; this is with a PREEMPT kernel so the tar goes to completion
without locking the system.

I started using a memory balloon to clear out the page cache.

procs -----------memory---------- ---swap-- -----io---- --system-- -----cpu-----
r b swpd free buff cache si so bi bo in cs us sy id wa st
0 0 98708 1676100 4112 39556 0 45 1566 1879 392 168 14 17 59 10 0
0 0 98708 1676100 4112 39604 0 0 0 0 37 35 0 0 100 0 0
0 0 98708 1676100 4112 39652 0 0 0 0 31 35 0 0 100 0 0
0 0 98708 1676100 4112 39652 0 0 0 0 26 27 0 0 100 0 0
0 0 98708 1676100 4112 39652 32 0 32 0 50 48 0 0 100 0 0
1 1 98708 1673984 4616 40576 192 0 1812 0 286 117 0 0 92 8 0
0 1 98708 1590540 4792 124236 0 0 41972 0 901 822 0 4 67 29 0
0 1 98708 1499524 4808 214820 0 0 45328 0 917 890 0 4 77 20 0
0 1 98708 1404532 4808 308316 64 0 46784 0 965 1033 0 3 77 20 0
0 1 98708 1312020 5612 399008 0 0 44836 0 1274 1075 0 6 77 18 0
0 1 98708 1304540 5628 406324 0 0 3600 47260 289 115 0 2 68 30 0
0 1 98708 1230760 5672 479020 0 0 36260 712 914 854 0 3 79 18 0
0 1 98708 1208696 5676 500016 0 0 10500 93888 421 251 0 3 67 31 0
0 1 98708 1157632 5676 551264 0 0 25600 0 658 518 0 2 75 23 0
0 1 98708 1052356 5680 655344 0 0 51972 0 1112 1112 0 4 78 19 0
0 2 98708 1024960 5680 679080 0 0 11904 87304 477 277 0 3 54 43 0
1 1 98708 1027288 5684 679432 0 0 128 2896 195 42 0 0 85 14 0
0 1 98708 947124 5688 758528 0 0 39552 4 932 899 0 4 68 28 0
0 1 98708 849792 5692 854272 0 0 47876 4468 1045 936 0 4 77 20 0
0 1 98708 813000 5820 890492 0 0 18176 18300 512 386 0 3 66 32 0
1 2 98708 794304 5820 908132 0 0 8832 95640 379 324 0 2 77 21 0
procs -----------memory---------- ---swap-- -----io---- --system-- -----cpu-----
r b swpd free buff cache si so bi bo in cs us sy id wa st
1 1 98708 794272 5824 908128 0 0 0 39456 212 58 0 1 73 26 0
0 1 98708 792344 5832 910452 0 0 1156 44440 193 62 0 1 74 25 0
0 1 98708 764392 5836 938288 0 0 13956 0 441 374 0 1 79 19 0
0 1 98708 660828 5840 1040232 0 0 50948 0 1141 992 0 6 65 29 0
0 1 98708 555156 5840 1144216 0 0 51968 0 1072 1008 0 4 76 20 0
0 1 98708 458692 5856 1239212 0 0 47496 120 995 1046 0 3 78 19 0
0 1 98708 411336 5856 1285896 0 0 23296 17764 536 463 0 3 71 26 0
1 1 98708 367716 5856 1328904 0 0 21504 22272 541 463 0 3 78 20 0
0 1 98708 330048 5864 1366044 0 0 18568 17524 529 515 0 2 78 20 0
0 1 98708 301620 5864 1393912 0 0 13952 28148 382 309 0 2 68 30 0
1 1 98708 245140 5872 1449576 0 0 27776 13224 610 567 0 3 73 24 0
0 2 98708 226868 5872 1467000 0 0 8704 114804 375 322 0 2 77 21 0
0 2 98708 226268 5876 1467552 0 0 260 51720 204 40 0 1 75 24 0
0 1 98708 226548 5880 1467536 0 0 4 13180 149 29 0 1 70 30 0
2 0 98708 159792 5880 1533844 0 0 33152 0 721 747 0 2 81 17 0
0 1 98708 55252 5888 1636692 0 0 51456 104 1021 977 0 5 70 25 0
1 1 98708 22804 2680 1671344 0 0 53632 0 1806 1188 0 20 59 21 0
2 1 98708 22832 2672 1671372 0 0 40704 11196 1967 1126 0 22 60 17 0
1 1 98708 22908 2676 1671196 0 0 24580 23040 1621 705 0 35 36 30 0
1 1 98708 22968 2676 1670796 0 0 21504 22212 1576 674 0 22 60 18 0
1 1 98708 22672 2684 1670996 0 0 17920 16968 1450 687 0 27 51 23 0
procs -----------memory---------- ---swap-- -----io---- --system-- -----cpu-----
r b swpd free buff cache si so bi bo in cs us sy id wa st
2 1 98708 22904 2452 1671344 0 0 24832 25780 1632 714 0 27 50 23 0
1 2 98708 22764 2452 1670724 0 0 15488 96304 1459 531 0 37 33 30 0
2 2 98708 22640 2452 1670724 0 0 0 60300 1219 356 0 20 45 35 0
1 1 98708 22744 2452 1670688 0 0 768 29312 1200 253 0 29 36 34 0
2 0 98708 22964 2460 1671040 0 0 10752 68 1345 445 0 23 56 21 0
1 1 98708 22652 2536 1671560 0 0 51540 0 2113 1246 0 28 49 23 0
1 1 98708 22676 2528 1671768 0 0 118276 14604 4875 2681 0 27 51 22 0
1 1 98708 22644 2528 1671804 0 0 23680 18172 1539 774 0 29 44 27 0
1 1 98708 22720 2536 1671464 0 0 21888 24664 1494 651 0 24 55 22 0
2 1 98708 22672 2496 1671592 0 0 22528 18144 1493 643 0 30 42 28 0
1 1 98708 22732 2488 1671508 0 0 22528 25192 1537 726 0 23 55 21 0
1 2 98708 22884 2488 1670504 0 0 18816 89440 1464 581 0 31 37 32 0
1 2 98708 22740 2488 1670764 0 0 768 75012 1196 254 0 22 55 23 0
1 2 98708 22848 2492 1670600 0 0 256 15512 1225 358 0 29 27 44 0
1 1 98708 22620 2500 1671628 0 0 19076 11408 1490 575 0 23 54 23 0
1 1 98708 22988 2500 1671108 0 0 55040 0 2118 1203 0 27 51 22 0
1 1 98708 22876 2488 1671044 0 0 53760 0 2138 1323 0 28 49 23 0
1 1 98708 22904 2500 1671012 0 0 48012 5528 2049 1097 0 32 43 25 0
1 1 98708 22856 2508 1670876 0 0 21504 18536 1500 646 0 23 55 22 0
2 0 98708 22684 2504 1670972 0 0 23680 23552 1537 784 0 29 44 26 0
1 1 98708 22960 2500 1670520 0 0 23808 25552 1529 677 0 24 55 21 0
procs -----------memory---------- ---swap-- -----io---- --system-- -----cpu-----
r b swpd free buff cache si so bi bo in cs us sy id wa st
2 1 98708 22820 2504 1670804 0 0 21764 25104 1507 647 0 30 43 26 0
1 1 98708 22880 2504 1670812 0 0 24192 18900 1541 768 0 23 55 22 0
2 3 98708 22716 2512 1670068 0 0 2816 132592 1257 303 0 30 30 40 0
1 1 98708 22944 2512 1669940 0 0 512 40400 1165 247 0 23 57 20 0
1 1 98708 22672 2512 1670624 0 0 512 0 1186 349 0 28 44 29 0
1 1 98708 22788 2504 1670416 0 0 51200 0 2130 1158 0 25 53 22 0
1 1 98708 22660 2508 1670688 0 0 50564 0 1972 1161 0 31 43 26 0
1 1 98708 22780 2516 1670088 0 0 50176 64 2028 1204 0 24 56 20 0
1 1 98708 22856 2508 1669884 0 0 32640 19040 1723 819 0 31 43 26 0
3 0 98708 22808 2512 1669812 0 0 23812 19176 1535 670 0 23 55 22 0
1 1 98708 22744 2504 1669728 0 0 21888 24872 1535 653 0 30 43 27 0
1 1 98708 22696 2504 1669636 0 0 22784 25032 1529 647 0 23 55 21 0
1 1 98708 22632 2512 1669596 0 0 21896 19228 1515 647 0 30 43 27 0
1 2 98708 22676 2512 1668992 0 0 9600 74484 1366 418 0 24 43 33 0
1 2 98708 22780 2512 1668772 0 0 256 107120 1211 245 0 29 42 30 0
1 1 98708 23196 2512 1668920 0 0 256 40 1158 228 0 23 57 20 0
2 1 98708 22968 2512 1669092 0 0 47360 0 2023 1097 0 26 49 25 0
1 1 98708 22864 2704 1668784 0 0 39992 116 1878 994 0 27 49 24 0
1 1 98708 22704 2708 1668872 0 0 53388 0 2166 1206 0 28 49 23 0
1 1 98708 22796 2712 1668472 0 0 39940 12784 1873 966 0 27 52 21 0
2 1 98708 22624 2712 1668508 0 0 21888 21904 1498 645 0 30 43 27 0
procs -----------memory---------- ---swap-- -----io---- --system-- -----cpu-----
r b swpd free buff cache si so bi bo in cs us sy id wa st
2 1 98708 22700 2704 1668688 0 0 24320 26240 1548 685 0 23 55 22 0
1 1 98708 22776 2712 1668284 0 0 18048 17532 1438 586 0 30 43 27 0
1 1 98708 22868 2712 1668336 0 0 24192 21764 1552 661 0 23 55 23 0
2 2 98708 22772 2716 1667696 0 0 12420 88772 1433 494 0 30 41 28 0
1 2 98708 23424 2720 1667584 0 0 4 9092 1181 246 0 23 40 37 0
1 1 98708 23120 2724 1667584 0 0 4 86380 1262 269 0 29 42 29 0
1 0 98708 23408 2984 1667708 0 0 364 0 1187 249 0 22 60 18 0
2 0 98708 23416 2992 1667772 0 0 0 120 1045 246 0 25 74 1 0
1 0 98708 23408 2992 1667720 0 0 0 0 1025 221 0 26 74 0 0
1 0 98708 23408 2992 1667720 0 0 0 0 1030 225 0 28 72 0 0
1 0 98708 23408 2992 1667720 0 0 0 0 1024 224 0 23 77 0 0
1 0 98708 23408 2992 1667720 0 0 0 0 1034 226 0 28 72 0 0
1 0 98708 23408 2992 1667720 0 0 0 0 1022 221 0 23 77 0 0
1 0 98708 23408 2992 1667720 0 0 0 0 1032 227 0 28 72 0 0
1 0 98708 23408 2992 1667720 0 0 0 0 1024 221 0 23 77 0 0
1 0 98708 23408 2992 1667720 0 0 0 0 1034 228 0 28 72 0 0

kswapd goes up to 99% at nearly the end of the second page. I ^C on the
tar but kswapd stays at 99% for the end.

James


--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/