== replay-dual test 26: dbench and tar with mds failover ========================================================== 13:09:57 (1789405797) Starting client oleg307-client.virtnet: -o user_xattr,flock 192.168.203.107@tcp:/lustre /mnt/lustre Started clients oleg307-client.virtnet: 192.168.203.107@tcp:/lustre on /mnt/lustre type lustre (rw,flock,checksum,encrypt,lazystatfs,lruresize,nolock,statfs_project,nosync_on_close,nouser_fid2path,user_xattr,verbose) Started tar loop with pid 45102 Started dbench loop with 45103 looking for dbench program /usr/bin/dbench found dbench client file /usr/share/dbench/client.txt UUID 1K-blocks Used Available Use% Mounted on lustre-MDT0000_UUID 2210560 2048 2206464 1% /mnt/lustre[MDT:0] lustre-OST0000_UUID 3771392 2048 3762176 1% /mnt/lustre[OST:0] lustre-OST0001_UUID 3771392 18432 3734528 1% /mnt/lustre[OST:1] filesystem_summary: 7542784 20480 7496704 1% /mnt/lustre '/usr/share/dbench/client.txt' -> 'client.txt' running 'dbench 1 -D /mnt/lustre2/d26.replay-dual/run_dbench -t 100' on /mnt/lustre at Mon Sep 14 13:10:02 EDT 2026 waiting for dbench pid 45170 dbench version 4.00 - Copyright Andrew Tridgell 1999-2004 Running for 100 seconds with load 'client.txt' and minimum warmup 20 secs failed to create barrier semaphore 0 of 1 processes prepared for launch 0 sec 1 of 1 processes prepared for launch 0 sec releasing clients 1 31 1.05 MB/sec warmup 1 sec latency 73.756 ms 1 88 2.03 MB/sec warmup 2 sec latency 139.154 ms 1 122 1.83 MB/sec warmup 3 sec latency 182.521 ms 1 161 1.65 MB/sec warmup 4 sec latency 116.462 ms test_26 fail mds1 1 times 1 190 1.51 MB/sec warmup 5 sec latency 189.435 ms 1 219 1.40 MB/sec warmup 6 sec latency 178.067 ms Failing mds1 on oleg307-server 1 255 1.38 MB/sec warmup 7 sec latency 141.571 ms 1 287 1.33 MB/sec warmup 8 sec latency 145.916 ms Stopping /mnt/lustre-mds1 (opts:) on oleg307-server 1 315 1.27 MB/sec warmup 9 sec latency 188.079 ms 1 323 1.17 MB/sec warmup 10 sec latency 753.245 ms 1 323 1.07 MB/sec warmup 11 sec latency 1753.492 ms 1 323 0.98 MB/sec warmup 12 sec latency 2756.226 ms 13:10:15 (1789405815) shut down facet: mds1 facet_host: oleg307-server facet_failover_host: oleg307-server 1 323 0.90 MB/sec warmup 13 sec latency 3758.388 ms 1 323 0.84 MB/sec warmup 14 sec latency 4758.745 ms 1 323 0.78 MB/sec warmup 15 sec latency 5759.387 ms 1 323 0.73 MB/sec warmup 16 sec latency 6759.801 ms 1 323 0.69 MB/sec warmup 17 sec latency 7761.637 ms 1 323 0.65 MB/sec warmup 18 sec latency 8762.053 ms 1 323 0.62 MB/sec warmup 19 sec latency 9762.486 ms 1 323 0.00 MB/sec execute 1 sec latency 11763.203 ms 1 323 0.00 MB/sec execute 2 sec latency 12763.518 ms Failover mds1 to oleg307-server mount facets: mds1 1 323 0.00 MB/sec execute 3 sec latency 13770.861 ms 1 323 0.00 MB/sec execute 4 sec latency 14771.047 ms 1 323 0.00 MB/sec execute 5 sec latency 15772.135 ms 1 323 0.00 MB/sec execute 6 sec latency 16774.587 ms 1 323 0.00 MB/sec execute 7 sec latency 17774.798 ms Start mds1: mount -t lustre -o localrecov lustre-mdt1/mdt1 /mnt/lustre-mds1 1 323 0.00 MB/sec execute 8 sec latency 18775.030 ms 1 323 0.00 MB/sec execute 9 sec latency 19775.461 ms 1 323 0.00 MB/sec execute 10 sec latency 20775.871 ms 1 323 0.00 MB/sec execute 11 sec latency 21776.194 ms 1 323 0.00 MB/sec execute 12 sec latency 22776.413 ms 1 323 0.00 MB/sec execute 13 sec latency 23776.639 ms 1 323 0.00 MB/sec execute 14 sec latency 24776.902 ms 1 323 0.00 MB/sec execute 15 sec latency 25778.824 ms 1 323 0.00 MB/sec execute 16 sec latency 26779.023 ms 1 323 0.00 MB/sec execute 17 sec latency 27779.299 ms 1 323 0.00 MB/sec execute 18 sec latency 28781.076 ms 1 323 0.00 MB/sec execute 19 sec latency 29781.659 ms 1 323 0.00 MB/sec execute 20 sec latency 30782.042 ms oleg307-server: oleg307-server.virtnet: executing set_default_debug -1 all 1 323 0.00 MB/sec execute 21 sec latency 31782.335 ms pdsh@oleg307-client: oleg307-server: ssh exited with exit code 1 1 323 0.00 MB/sec execute 22 sec latency 32783.287 ms Started lustre-MDT0000 13:10:45 (1789405845) targets are mounted 13:10:45 (1789405845) facet_failover done 1 323 0.00 MB/sec execute 23 sec latency 33783.629 ms 1 323 0.00 MB/sec execute 24 sec latency 34784.394 ms 1 323 0.00 MB/sec execute 25 sec latency 35784.588 ms 1 323 0.00 MB/sec execute 26 sec latency 36786.472 ms 1 323 0.00 MB/sec execute 27 sec latency 37787.810 ms 1 323 0.00 MB/sec execute 28 sec latency 38789.028 ms 1 347 0.02 MB/sec execute 29 sec latency 39012.222 ms 1 382 0.06 MB/sec execute 30 sec latency 136.090 ms 1 421 0.10 MB/sec execute 31 sec latency 130.562 ms 1 461 0.14 MB/sec execute 32 sec latency 158.087 ms 1 501 0.17 MB/sec execute 33 sec latency 107.997 ms oleg307-client.virtnet: executing wait_import_state_mount (FULL|IDLE) mdc.lustre-MDT0000-mdc-*.mds_server_uuid 1475 0 1 539 0.21 MB/sec execute 34 sec latency 153.383 ms 1 615 0.26 MB/sec execute 35 sec latency 140.577 ms mdc.lustre-MDT0000-mdc-*.mds_server_uuid in FULL state after 0 sec 1 664 0.26 MB/sec execute 36 sec latency 229.077 ms 1 699 0.26 MB/sec execute 37 sec latency 245.039 ms 1 730 0.25 MB/sec execute 38 sec latency 184.463 ms 1 769 0.25 MB/sec execute 39 sec latency 110.502 ms 1 798 0.24 MB/sec execute 40 sec latency 201.580 ms 1 851 0.24 MB/sec execute 41 sec latency 97.186 ms UUID 1K-blocks Used Available Use% Mounted on lustre-MDT0000_UUID 2210560 2304 2206208 1% /mnt/lustre[MDT:0] lustre-OST0000_UUID 3771392 11264 3732480 1% /mnt/lustre[OST:0] lustre-OST0001_UUID 3771392 35840 3657728 1% /mnt/lustre[OST:1] filesystem_summary: 7542784 47104 7390208 1% /mnt/lustre 1 878 0.24 MB/sec execute 42 sec latency 237.383 ms 1 913 0.23 MB/sec execute 43 sec latency 148.377 ms 1 949 0.23 MB/sec execute 44 sec latency 290.716 ms 1 985 0.23 MB/sec execute 45 sec latency 134.626 ms 1 1067 0.24 MB/sec execute 46 sec latency 73.831 ms 1 1136 0.24 MB/sec execute 47 sec latency 79.701 ms test_26 fail mds1 2 times 1 1192 0.24 MB/sec execute 48 sec latency 57.110 ms Failing mds1 on oleg307-server 1 1244 0.23 MB/sec execute 49 sec latency 64.285 ms Stopping /mnt/lustre-mds1 (opts:) on oleg307-server 1 1394 0.29 MB/sec execute 50 sec latency 85.619 ms 1 1409 0.28 MB/sec execute 51 sec latency 600.385 ms 1 1409 0.27 MB/sec execute 52 sec latency 1602.524 ms 13:11:15 (1789405875) shut down facet: mds1 facet_host: oleg307-server facet_failover_host: oleg307-server 1 1409 0.27 MB/sec execute 53 sec latency 2602.767 ms 1 1409 0.26 MB/sec execute 54 sec latency 3603.742 ms 1 1409 0.26 MB/sec execute 55 sec latency 4605.851 ms 1 1409 0.25 MB/sec execute 56 sec latency 5606.079 ms 1 1409 0.25 MB/sec execute 57 sec latency 6606.300 ms 1 1409 0.25 MB/sec execute 58 sec latency 7606.531 ms 1 1409 0.24 MB/sec execute 59 sec latency 8606.745 ms 1 1409 0.24 MB/sec execute 60 sec latency 9606.928 ms 1 1409 0.23 MB/sec execute 61 sec latency 10607.765 ms 1 1409 0.23 MB/sec execute 62 sec latency 11611.508 ms Failover mds1 to oleg307-server 1 1409 0.23 MB/sec execute 63 sec latency 12611.873 ms mount facets: mds1 1 1409 0.22 MB/sec execute 64 sec latency 13612.033 ms 1 1409 0.22 MB/sec execute 65 sec latency 14612.708 ms 1 1409 0.22 MB/sec execute 66 sec latency 15616.183 ms 1 1409 0.21 MB/sec execute 67 sec latency 16617.161 ms Start mds1: mount -t lustre -o localrecov lustre-mdt1/mdt1 /mnt/lustre-mds1 1 1409 0.21 MB/sec execute 68 sec latency 17617.478 ms 1 1409 0.21 MB/sec execute 69 sec latency 18617.645 ms 1 1409 0.20 MB/sec execute 70 sec latency 19620.429 ms 1 1409 0.20 MB/sec execute 71 sec latency 20620.678 ms 1 1409 0.20 MB/sec execute 72 sec latency 21621.895 ms oleg307-server: oleg307-server.virtnet: executing set_default_debug -1 all 1 1409 0.20 MB/sec execute 73 sec latency 22622.127 ms pdsh@oleg307-client: oleg307-server: ssh exited with exit code 1 1 1409 0.19 MB/sec execute 74 sec latency 23627.783 ms Started lustre-MDT0000 13:11:37 (1789405897) targets are mounted 13:11:37 (1789405897) facet_failover done 1 1426 0.19 MB/sec execute 75 sec latency 24160.456 ms 1 1492 0.19 MB/sec execute 76 sec latency 201.562 ms 1 1558 0.19 MB/sec execute 77 sec latency 72.377 ms 1 1605 0.18 MB/sec execute 78 sec latency 95.735 ms 1 1644 0.18 MB/sec execute 79 sec latency 253.458 ms oleg307-client.virtnet: executing wait_import_state_mount (FULL|IDLE) mdc.lustre-MDT0000-mdc-*.mds_server_uuid 1475 0 1 1686 0.18 MB/sec execute 80 sec latency 132.152 ms 1 1729 0.18 MB/sec execute 81 sec latency 130.776 ms mdc.lustre-MDT0000-mdc-*.mds_server_uuid in FULL state after 0 sec 1 1802 0.18 MB/sec execute 82 sec latency 70.236 ms 1 1851 0.18 MB/sec execute 83 sec latency 83.389 ms 1 1888 0.17 MB/sec execute 84 sec latency 245.503 ms 1 1978 0.17 MB/sec execute 85 sec latency 121.734 ms 1 2018 0.17 MB/sec execute 86 sec latency 73.249 ms 1 2088 0.17 MB/sec execute 87 sec latency 83.969 ms UUID 1K-blocks Used Available Use% Mounted on lustre-MDT0000_UUID 2210560 2688 2205824 1% /mnt/lustre[MDT:0] lustre-OST0000_UUID 3771392 15360 3709952 1% /mnt/lustre[OST:0] lustre-OST0001_UUID 3771392 57344 3676160 2% /mnt/lustre[OST:1] filesystem_summary: 7542784 72704 7386112 1% /mnt/lustre 1 2150 0.17 MB/sec execute 88 sec latency 91.394 ms 1 2185 0.17 MB/sec execute 89 sec latency 110.880 ms 1 2247 0.17 MB/sec execute 90 sec latency 118.371 ms 1 2342 0.18 MB/sec execute 91 sec latency 49.957 ms 1 2377 0.18 MB/sec execute 92 sec latency 230.379 ms 1 2416 0.18 MB/sec execute 93 sec latency 145.753 ms test_26 fail mds1 3 times 1 2442 0.17 MB/sec execute 94 sec latency 212.621 ms 1 2477 0.18 MB/sec execute 95 sec latency 247.992 ms Failing mds1 on oleg307-server 1 2555 0.18 MB/sec execute 96 sec latency 301.844 ms Stopping /mnt/lustre-mds1 (opts:) on oleg307-server 1 2566 0.18 MB/sec execute 97 sec latency 686.670 ms 1 2567 0.17 MB/sec execute 98 sec latency 1322.010 ms 1 2567 0.17 MB/sec execute 99 sec latency 1369.165 ms 13:12:02 (1789405922) shut down 1 cleanup 100 sec facet: mds1 facet_host: oleg307-server facet_failover_host: oleg307-server 1 cleanup 101 sec 1 cleanup 102 sec 1 cleanup 103 sec 1 cleanup 104 sec 1 cleanup 105 sec 1 cleanup 106 sec 1 cleanup 107 sec 1 cleanup 108 sec 1 cleanup 109 sec 1 cleanup 110 sec Failover mds1 to oleg307-server mount facets: mds1 1 cleanup 111 sec 1 cleanup 112 sec 1 cleanup 113 sec 1 cleanup 114 sec 1 cleanup 115 sec Start mds1: mount -t lustre -o localrecov lustre-mdt1/mdt1 /mnt/lustre-mds1 1 cleanup 116 sec 1 cleanup 117 sec 1 cleanup 118 sec 1 cleanup 119 sec 1 cleanup 120 sec 1 cleanup 121 sec 1 cleanup 122 sec 1 cleanup 123 sec 1 cleanup 124 sec 1 cleanup 125 sec 1 cleanup 126 sec 1 cleanup 127 sec 1 cleanup 128 sec 1 cleanup 129 sec oleg307-server: oleg307-server.virtnet: executing set_default_debug -1 all 1 cleanup 130 sec 1 cleanup 131 sec pdsh@oleg307-client: oleg307-server: ssh exited with exit code 1 1 cleanup 132 sec Started lustre-MDT0000 13:12:35 (1789405955) targets are mounted 13:12:35 (1789405955) facet_failover done 1 cleanup 133 sec 1 cleanup 134 sec 1 cleanup 135 sec 1 cleanup 136 sec 1 cleanup 137 sec 1 cleanup 138 sec 1 cleanup 139 sec 1 cleanup 140 sec 1 cleanup 141 sec 1 cleanup 142 sec 1 cleanup 143 sec 1 cleanup 144 sec 0 cleanup 144 sec Operation Count AvgLat MaxLat ---------------------------------------- NTCreateX 376 157.553 39012.203 Close 325 85.774 24160.291 Rename 15 82.951 212.589 Unlink 38 25.372 52.180 Qpathinfo 340 13.528 79.707 Qfileinfo 103 1.727 21.488 Qfsinfo 43 2.935 12.577 Sfileinfo 49 915.173 42506.480 Find 116 5.495 86.110 WriteX 379 12.025 78.506 ReadX 427 1.720 70.537 Flush 34 139.522 301.828 Throughput 0.173147 MB/sec 1 clients 1 procs max_latency=39012.222 ms stopping dbench on /mnt/lustre at Mon Sep 14 13:12:47 EDT 2026 with return code 0 clean dbench files on /mnt/lustre /mnt/lustre /mnt/lustre oleg307-client.virtnet: executing wait_import_state_mount (FULL|IDLE) mdc.lustre-MDT0000-mdc-*.mds_server_uuid 1475 0 removed 'client.txt' /mnt/lustre dbench successfully finished mdc.lustre-MDT0000-mdc-*.mds_server_uuid in FULL state after 0 sec looking for dbench program /usr/bin/dbench found dbench client file /usr/share/dbench/client.txt '/usr/share/dbench/client.txt' -> 'client.txt' running 'dbench 1 -D /mnt/lustre2/d26.replay-dual/run_dbench -t 100' on /mnt/lustre at Mon Sep 14 13:12:53 EDT 2026 waiting for dbench pid 48101 dbench version 4.00 - Copyright Andrew Tridgell 1999-2004 Running for 100 seconds with load 'client.txt' and minimum warmup 20 secs 0 of 1 processes prepared for launch 0 sec 1 of 1 processes prepared for launch 0 sec releasing clients 1 32 1.05 MB/sec warmup 1 sec latency 128.912 ms 1 113 2.62 MB/sec warmup 2 sec latency 133.749 ms 1 157 2.18 MB/sec warmup 3 sec latency 132.766 ms UUID 1K-blocks Used Available Use% Mounted on lustre-MDT0000_UUID 2210560 3200 2205312 1% /mnt/lustre[MDT:0] lustre-OST0000_UUID 3771392 28672 3728384 1% /mnt/lustre[OST:0] lustre-OST0001_UUID 3771392 25600 3741696 1% /mnt/lustre[OST:1] filesystem_summary: 7542784 54272 7470080 1% /mnt/lustre 1 193 1.91 MB/sec warmup 4 sec latency 190.646 ms 1 231 1.78 MB/sec warmup 5 sec latency 123.483 ms 1 271 1.70 MB/sec warmup 6 sec latency 118.109 ms 1 309 1.61 MB/sec warmup 7 sec latency 127.488 ms 1 349 1.56 MB/sec warmup 8 sec latency 136.618 ms test_26 fail mds1 4 times 1 389 1.53 MB/sec warmup 9 sec latency 124.390 ms 1 428 1.51 MB/sec warmup 10 sec latency 121.821 ms 1 465 1.47 MB/sec warmup 11 sec latency 124.510 ms Failing mds1 on oleg307-server 1 501 1.45 MB/sec warmup 12 sec latency 162.627 ms Stopping /mnt/lustre-mds1 (opts:) on oleg307-server 1 560 1.47 MB/sec warmup 13 sec latency 129.206 ms 1 600 1.50 MB/sec warmup 14 sec latency 605.104 ms 1 600 1.40 MB/sec warmup 15 sec latency 1605.387 ms 1 600 1.31 MB/sec warmup 16 sec latency 2605.907 ms 13:13:10 (1789405990) shut down facet: mds1 facet_host: oleg307-server facet_failover_host: oleg307-server 1 600 1.23 MB/sec warmup 17 sec latency 3609.351 ms 1 600 1.16 MB/sec warmup 18 sec latency 4611.052 ms 1 600 1.10 MB/sec warmup 19 sec latency 5612.438 ms 1 600 0.00 MB/sec execute 1 sec latency 7612.788 ms 1 600 0.00 MB/sec execute 2 sec latency 8612.995 ms 1 600 0.00 MB/sec execute 3 sec latency 9613.339 ms 1 600 0.00 MB/sec execute 4 sec latency 10614.109 ms 1 600 0.00 MB/sec execute 5 sec latency 11614.868 ms 1 600 0.00 MB/sec execute 6 sec latency 12615.259 ms Failover mds1 to oleg307-server mount facets: mds1 1 600 0.00 MB/sec execute 7 sec latency 13615.553 ms 1 600 0.00 MB/sec execute 8 sec latency 14615.784 ms 1 600 0.00 MB/sec execute 9 sec latency 15617.310 ms 1 600 0.00 MB/sec execute 10 sec latency 16617.762 ms 1 600 0.00 MB/sec execute 11 sec latency 17617.908 ms Start mds1: mount -t lustre -o localrecov lustre-mdt1/mdt1 /mnt/lustre-mds1 1 600 0.00 MB/sec execute 12 sec latency 18618.385 ms 1 600 0.00 MB/sec execute 13 sec latency 19618.816 ms 1 600 0.00 MB/sec execute 14 sec latency 20623.437 ms 1 600 0.00 MB/sec execute 15 sec latency 21623.820 ms 1 600 0.00 MB/sec execute 16 sec latency 22624.015 ms 1 600 0.00 MB/sec execute 17 sec latency 23624.275 ms 1 600 0.00 MB/sec execute 18 sec latency 24624.554 ms 1 600 0.00 MB/sec execute 19 sec latency 25624.998 ms 1 600 0.00 MB/sec execute 20 sec latency 26626.104 ms 1 600 0.00 MB/sec execute 21 sec latency 27626.427 ms 1 600 0.00 MB/sec execute 22 sec latency 28627.382 ms 1 600 0.00 MB/sec execute 23 sec latency 29629.573 ms 1 600 0.00 MB/sec execute 24 sec latency 30630.556 ms 1 600 0.00 MB/sec execute 25 sec latency 31630.761 ms oleg307-server: oleg307-server.virtnet: executing set_default_debug -1 all 1 600 0.00 MB/sec execute 26 sec latency 32631.013 ms pdsh@oleg307-client: oleg307-server: ssh exited with exit code 1 1 600 0.00 MB/sec execute 27 sec latency 33631.152 ms Started lustre-MDT0000 13:13:41 (1789406021) targets are mounted 13:13:41 (1789406021) facet_failover done 1 600 0.00 MB/sec execute 28 sec latency 34631.662 ms 1 600 0.00 MB/sec execute 29 sec latency 35638.673 ms 1 600 0.00 MB/sec execute 30 sec latency 36639.198 ms 1 600 0.00 MB/sec execute 31 sec latency 37639.881 ms 1 600 0.00 MB/sec execute 32 sec latency 38640.074 ms 1 600 0.00 MB/sec execute 33 sec latency 39640.307 ms 1 600 0.00 MB/sec execute 34 sec latency 40640.936 ms 1 636 0.00 MB/sec execute 35 sec latency 41128.573 ms 1 669 0.01 MB/sec execute 36 sec latency 329.113 ms 1 699 0.01 MB/sec execute 37 sec latency 140.176 ms 1 723 0.01 MB/sec execute 38 sec latency 255.698 ms oleg307-client.virtnet: executing wait_import_state_mount (FULL|IDLE) mdc.lustre-MDT0000-mdc-*.mds_server_uuid 1475 0 1 747 0.01 MB/sec execute 39 sec latency 125.073 ms 1 789 0.01 MB/sec execute 40 sec latency 197.578 ms mdc.lustre-MDT0000-mdc-*.mds_server_uuid in FULL state after 0 sec 1 808 0.01 MB/sec execute 41 sec latency 220.762 ms 1 865 0.01 MB/sec execute 42 sec latency 120.098 ms 1 890 0.02 MB/sec execute 43 sec latency 267.876 ms 1 916 0.02 MB/sec execute 44 sec latency 222.251 ms tar: Unexpected EOF in archive tar: Unexpected EOF in archive 1 969 0.02 MB/sec execute 45 sec latency 110.957 ms 1 995 0.02 MB/sec execute 46 sec latency 304.524 ms 1 1050 0.04 MB/sec execute 47 sec latency 185.125 ms 1 1122 0.04 MB/sec execute 48 sec latency 87.101 ms 1 1172 0.04 MB/sec execute 49 sec latency 97.002 ms 1 1223 0.04 MB/sec execute 50 sec latency 100.112 ms 1 1251 0.04 MB/sec execute 51 sec latency 272.056 ms 1 1384 0.10 MB/sec execute 52 sec latency 91.733 ms 1 1426 0.10 MB/sec execute 53 sec latency 106.639 ms 1 1481 0.09 MB/sec execute 54 sec latency 104.186 ms tar: Error is not recoverable: exiting now dbench killed by signal 15 stopping dbench on /mnt/lustre at Mon Sep 14 13:14:08 EDT 2026 with return code 0 48101 pts/0 S+ 0:00 dbench -c client.txt 1 -D /mnt/lustre2/d26.replay-dual/run_dbench -t 100 killed dbench main pid 48101 clean dbench files on /mnt/lustre /mnt/lustre /mnt/lustre removed 'client.txt' /mnt/lustre dbench successfully finished