[LU-11430] sanity test 271d: too many arguments

• Status: Resolved
• Project: Lustre
• Affects Version/s: Lustre 2.12.0
• Fix Version/s: Lustre 2.12.0
• Type: Bug
• Priority: Minor
• Reporter: Jian Yu
• Assignee: Mikhail Pershin
• Resolution: Fixed
• Severity: 3

Description

sanity test 271d failed as follows in DNE test sessions:

== sanity test 271d: DoM: read on open (1K file in reply buffer) ===================================== 05:03:01 (1534136581)
CMD: trevis-7vm4 /usr/sbin/lctl get_param -n version 2>/dev/null ||
				/usr/sbin/lctl lustre_build_version 2>/dev/null ||
				/usr/sbin/lctl --version 2>/dev/null | cut -d' ' -f2
warning: '-M' deprecated, use '--mdt-index' or '-m' instead
1+0 records in
1+0 records out
1000 bytes (1.0 kB) copied, 0.000146744 s, 6.8 MB/s
1+0 records in
1+0 records out
1000 bytes (1.0 kB) copied, 0.00174574 s, 573 kB/s
Append to the same page
/usr/lib64/lustre/tests/sanity.sh: line 17034: [: too many arguments
/usr/lib64/lustre/tests/sanity.sh: line 17034: 6
1
1
1: syntax error in expression (error token is "1
1
1")

Maloo reports:
https://testing.whamcloud.com/test_sets/792d8c78-9ff2-11e8-8ee3-52540065bddc
https://testing.whamcloud.com/test_sets/5b537c1a-c0ec-11e8-a9d9-52540065bddc
https://testing.whamcloud.com/test_sets/794a6aa2-bed1-11e8-b748-52540065bddc

The failure is affecting patch testing on master branch.

Comments

[James Nunez (Inactive)]

We see this failure in sanity 271e for review-dne-* testing. For example https://testing.whamcloud.com/test_sets/8e26abfe-c8d4-11e8-b589-52540065bddc .

I think the issue is

17161         local num=$(lctl get_param -n mdc.*.stats |
17162                 awk '/ost_read/ {print $2}')
17163         local ra=$(lctl get_param -n mdc.*.stats |
17164                 awk '/req_active/ {print $2}')
17165         local rw=$(lctl get_param -n mdc.*.stats |
17166                 awk '/req_waittime/ {print $2}')
17167 
17168         [ -z $num ] || error "$num READ RPC occured"
17169         [ $ra == $rw ] || error "$((ra - rw)) resend occured"
17170         echo "... DONE"

For DNE, getting the param mdc.*.stats and awking for req_waittime and req_active will return multiple values, one for each mdc, and this can cause the error seen here.

I'm guessing we need to be more selective about what mdc to pull the stats from.

Mike - any thoughts on this failure?

[Mikhail Pershin]

yes, indeed, I will prepare patch ASAP

[Gerrit Updater]

Mike Pershin ([e-mail]) uploaded a new patch: https://review.whamcloud.com/33490
Subject: LU-11430 tests: get MDC stats by index
Project: fs/lustre-release
Branch: master
Current Patch Set: 1
Commit: d3f7d8f1ac35abe37cc86b7ef82e6e76ab1dd575

[Gerrit Updater]

Oleg Drokin ([e-mail]) merged in patch https://review.whamcloud.com/33490/
Subject: LU-11430 tests: get MDC stats by index
Project: fs/lustre-release
Branch: master
Current Patch Set:
Commit: c50bd08b35757d20e606c10bb5b4474011dd07f6