[LU-3419] lustre-rsync-test test_2b: dbench not running when killed

• Status: Resolved
• Project: Lustre
• Affects Version/s: Lustre 2.4.0
• Type: Bug
• Priority: Minor
• Reporter: Maloo
• Assignee: WC Triage
• Resolution: Cannot Reproduce
• Labels: zfs
• Severity: 3

Description

This issue was created by maloo for sarah <[e-mail]>

This issue relates to the following test suite run: http://maloo.whamcloud.com/test_sets/8f8652ea-c797-11e2-9f90-52540035b04c.

The sub-test test_2b failed with the following error:

Failure in replication; differences found.

There were similar issues reported for interop testing, but this one is hit on zfs with both server and client run b2_4RC2

Comments

[Oleg Drokin]

seems dbench was really slow to start:

== lustre-rsync-test test 2b: Replicate files changed by dbench. == 03:18:10 (1369736290)
CMD: client-30vm7 lctl --device lustre-MDT0000 changelog_register -n
lustre-MDT0000: Registered changelog user cl3
CMD: client-30vm7 lctl get_param -n mdd.lustre-MDT0000.changelog_users
copying /usr/share/dbench/client.txt to /mnt/lustre/d0.lustre-rsync-test/d2/client.txt
copying necessary libs to /mnt/lustre/d0.lustre-rsync-test/d2
lib64/libpopt.so.0
lib64/libc.so.6
lib64/ld-linux-x86-64.so.2
PIDs:
Stopping dbench
usage: kill [ -s signal | -p ] [ -a ] pid ...
       kill -l [ signal ]
Starting replication
running 'dbench 2 -t 60' in /mnt/lustre/d0.lustre-rsync-test/d2 at Tue May 28 03:18:31 PDT 2013
dbench PID=20797
dbench version 4.00 - Copyright Andrew Tridgell 1999-2004

Running for 60 seconds with load 'client.txt' and minimum warmup 12 secs
0 of 2 processes prepared for launch   0 sec
2 of 2 processes prepared for launch   0 sec
releasing clients
   2       114    10.49 MB/sec  warmup   1 sec  latency 564.452 ms
   2       664    21.16 MB/sec  warmup   2 sec  latency 228.266 ms

so dbench was running in the middle of rsync.

I think we need to change the way we run dbench so that the function takes the time argument too?