[LU-8804] sanity test_101b: FAIL: Too many (1) discarded pages with size (8192) Created: 07/Nov/16  Updated: 06/Jul/21  Resolved: 06/Jul/21

Status: Resolved
Project: Lustre
Component/s: None
Affects Version/s: Lustre 2.9.0
Fix Version/s: None

Type: Bug Priority: Minor
Reporter: Steve Guminski (Inactive) Assignee: WC Triage
Resolution: Cannot Reproduce Votes: 0
Labels: None

Severity: 3
Rank (Obsolete): 9223372036854775807

 Description   

Sanity test 101b failed as follows:

== sanity test 101b: check stride-io mode read-ahead ================================================= 15:17:21 (1478272641)
CMD: trevis-3vm7 /usr/sbin/lctl set_param -n obdfilter.lustre-OST*.read_cache_enable=0 		osd-*.lustre-OST*.read_cache_enable=0 2>&1
CMD: trevis-3vm7 /usr/sbin/lctl set_param -n obdfilter.lustre-OST*.writethrough_cache_enable=0 		osd-*.lustre-OST*.writethrough_cache_enable=0 2>&1

1.498804s, 5.59687MB/s
llite.lustre-ffff880037b69800.read_ahead_stats=
snapshot_time             1478272647.905158 secs.usecs
hits                      1518 samples [pages]
misses                    530 samples [pages]
readpage not consecutive  8 samples [pages]
miss inside window        128 samples [pages]
failed grab_cache_page    256 samples [pages]
read but discarded        1 samples [pages]
zero size window          274 samples [pages]
 sanity test_101b: @@@@@@ FAIL: Too many (1) discarded pages with size (8192) 
  Trace dump:
  = /usr/lib64/lustre/tests/test-framework.sh:4841:error()
  = /usr/lib64/lustre/tests/sanity.sh:6579:ra_check_101()
  = /usr/lib64/lustre/tests/sanity.sh:6612:test_101b()
  = /usr/lib64/lustre/tests/test-framework.sh:5117:run_one()
  = /usr/lib64/lustre/tests/test-framework.sh:5156:run_one_logged()
  = /usr/lib64/lustre/tests/test-framework.sh:5003:run_test()
  = /usr/lib64/lustre/tests/sanity.sh:6617:main()
Dumping lctl log to /logdir/test_logs/2016-11-04/lustre-reviews-el7-x86_64--review-ldiskfs--1_10_1__42325__-69957394635600-134451/sanity.test_101b.*.1478272648.log
CMD: trevis-3vm12,trevis-3vm5.trevis.hpdd.intel.com,trevis-3vm6,trevis-3vm7 /usr/sbin/lctl dk > /logdir/test_logs/2016-11-04/lustre-reviews-el7-x86_64--review-ldiskfs--1_10_1__42325__-69957394635600-134451/sanity.test_101b.debug_log.\$(hostname -s).1478272648.log;
         dmesg > /logdir/test_logs/2016-11-04/lustre-reviews-el7-x86_64--review-ldiskfs--1_10_1__42325__-69957394635600-134451/sanity.test_101b.dmesg.\$(hostname -s).1478272648.log
Resetting fail_loc on all nodes...CMD: trevis-3vm12,trevis-3vm5.trevis.hpdd.intel.com,trevis-3vm6,trevis-3vm7 lctl set_param -n fail_loc=0 	    fail_val=0 2>/dev/null
done.
CMD: trevis-3vm7 /usr/sbin/lctl set_param -n obdfilter.lustre-OST*.read_cache_enable=1 		osd-*.lustre-OST*.read_cache_enable=1 2>&1

Maloo reports:
https://testing.hpdd.intel.com/test_sets/41939346-a2ad-11e6-8e31-5254006e85c2



 Comments   
Comment by Andreas Dilger [ 07/Nov/16 ]

Steve, for future bug reports, please include the Lustre version being tested. In this case, looking at the Maloo test logs I see it is 2.8.59.79.g233cb43, which would be reported as Affects Version: Lustre 2.9.0 here in Jira.

Generated at Sat Feb 10 02:20:41 UTC 2024 using Jira 9.4.14#940014-sha1:734e6822bbf0d45eff9af51f82432957f73aa32c.