Bug #11781
closedmultiple_rsync failure with fuse client
Added by John Spray almost 9 years ago. Updated almost 9 years ago.
0%
Description
The second rsync run is transferring some data, so either timestamps are going wrong, dentries are going missing, or the infrastructure is changing clocks beneath us.
http://pulpito.ceph.com/teuthology-2015-05-15_23:04:01-fs-master-testing-basic-multi/895391/
Updated by John Spray almost 9 years ago
- Project changed from devops to CephFS
- Category set to 45
Wrong project, sorry.
Updated by Greg Farnum almost 9 years ago
Possibly a dup of #9884? :/ I'm not sure how accurate that "too many files" is as the actual bug cause, though...
Updated by John Spray almost 9 years ago
#9884 was the case where it timed out b/c there were just so many files, whereas in this instance it's actually failing outright.
Updated by Zheng Yan almost 9 years ago
- Status changed from New to Rejected
2015-05-17T06:48:31.565 INFO:tasks.workunit.client.0.plana96.stdout:sent 6,261,568,409 bytes received 1,496,889 bytes 1,529,255.35 bytes/sec 2015-05-17T06:48:31.565 INFO:tasks.workunit.client.0.plana96.stdout:total size is 6,254,714,661 speedup is 1.00 2015-05-17T06:48:31.571 INFO:tasks.workunit.client.0.plana96.stderr:+ echo we should get 4 here if no additional files are transfered 2015-05-17T06:48:31.572 INFO:tasks.workunit.client.0.plana96.stdout:we should get 4 here if no additional files are transfered 2015-05-17T06:48:31.572 INFO:tasks.workunit.client.0.plana96.stderr:+ sudo rsync -auv --exclude local/ /usr/ usr.1 2015-05-17T06:48:31.572 INFO:tasks.workunit.client.0.plana96.stderr:+ tee /tmp/18207 2015-05-17T06:48:31.614 INFO:tasks.workunit.client.0.plana96.stdout:sending incremental file list 2015-05-17T06:50:15.753 INFO:tasks.workunit.client.0.plana96.stdout:share/doc/ 2015-05-17T06:58:08.209 INFO:tasks.workunit.client.0.plana96.stdout: 2015-05-17T06:58:08.209 INFO:tasks.workunit.client.0.plana96.stdout:sent 2,426,491 bytes received 11,316 bytes 4,221.31 bytes/sec 2015-05-17T06:58:08.209 INFO:tasks.workunit.client.0.plana96.stdout:total size is 6,254,711,077 speedup is 2,565.71 2015-05-17T06:58:08.213 INFO:tasks.workunit.client.0.plana96.stderr:+ hexdump -C /tmp/18207 2015-05-17T06:58:08.225 INFO:tasks.workunit.client.0.plana96.stdout:00000000 73 65 6e 64 69 6e 67 20 69 6e 63 72 65 6d 65 6e |sending incremen| 2015-05-17T06:58:08.225 INFO:tasks.workunit.client.0.plana96.stdout:00000010 74 61 6c 20 66 69 6c 65 20 6c 69 73 74 0a 73 68 |tal file list.sh| 2015-05-17T06:58:08.225 INFO:tasks.workunit.client.0.plana96.stdout:00000020 61 72 65 2f 64 6f 63 2f 0a 0a 73 65 6e 74 20 32 |are/doc/..sent 2| 2015-05-17T06:58:08.226 INFO:tasks.workunit.client.0.plana96.stdout:00000030 2c 34 32 36 2c 34 39 31 20 62 79 74 65 73 20 20 |,426,491 bytes | 2015-05-17T06:58:08.226 INFO:tasks.workunit.client.0.plana96.stdout:00000040 72 65 63 65 69 76 65 64 20 31 31 2c 33 31 36 20 |received 11,316 | 2015-05-17T06:58:08.226 INFO:tasks.workunit.client.0.plana96.stdout:00000050 62 79 74 65 73 20 20 34 2c 32 32 31 2e 33 31 20 |bytes 4,221.31 | 2015-05-17T06:58:08.226 INFO:tasks.workunit.client.0.plana96.stdout:00000060 62 79 74 65 73 2f 73 65 63 0a 74 6f 74 61 6c 20 |bytes/sec.total | 2015-05-17T06:58:08.227 INFO:tasks.workunit.client.0.plana96.stdout:00000070 73 69 7a 65 20 69 73 20 36 2c 32 35 34 2c 37 31 |size is 6,254,71| 2015-05-17T06:58:08.227 INFO:tasks.workunit.client.0.plana96.stdout:00000080 31 2c 30 37 37 20 20 73 70 65 65 64 75 70 20 69 |1,077 speedup i| 2015-05-17T06:58:08.227 INFO:tasks.workunit.client.0.plana96.stdout:00000090 73 20 32 2c 35 36 35 2e 37 31 0a |s 2,565.71.| 2015-05-17T06:58:08.227 INFO:tasks.workunit.client.0.plana96.stdout:0000009b
log for updating share/doc's mtime.
2015-05-17 04:51:01.413849 7f3f24ff1700 3 client.4125 ll_setattr 10000010209.head mask 18 2015-05-17 04:51:01.413864 7f3f24ff1700 10 client.4125 mark_caps_dirty 10000010209.head(ref=3 ll_ref=2123 cap_refs={} open={} mode=40755 size=0/0 mtime=2015-05-17 04:08:08.209612 caps=pAsLsXsFsx(0=pAsLsXsFsx) COMPLETE parents=0x7f3ea000c9b0 0x7f3f1ab80a50) - -> Fx 2015-05-17 04:51:01.413879 7f3f24ff1700 10 client.4125 fill_stat on 10000010209 snap/devhead mode 040755 mtime 2015-05-17 04:08:08.209612 ctime 2015-05-17 04:51:01.413864
2015-05-17 06:50:09.324997 7f3f257f2700 10 client.4125 send_request client_request(unknown.0:631558 setattr mtime=2015-05-17 06:48:08.028113 atime=2015-05-17 06:50:09.317534 #10000010209 2015-05-17 06:50:09.324974) v2 to mds.0
you can see that /usr/share/doc's mtime changed (by someone else).
Updated by John Spray almost 9 years ago
Hmm, maybe using /usr/ as a source of workload files is something we just need to stop doing, there's no particular reason to expect it to be constant
Updated by Greg Farnum almost 9 years ago
#11807. I'm not actually sure what good options there are for replacement, but it won't get lost.