[LU-14371] replay-single test 65b fails with 'No early reply' Created: 27/Jan/21 Updated: 07/Oct/22 Resolved: 07/Oct/22 |
|
| Status: | Resolved |
| Project: | Lustre |
| Component/s: | None |
| Affects Version/s: | Lustre 2.14.0 |
| Fix Version/s: | None |
| Type: | Bug | Priority: | Minor |
| Reporter: | James Nunez (Inactive) | Assignee: | WC Triage |
| Resolution: | Duplicate | Votes: | 0 |
| Labels: | None | ||
| Environment: |
(So far only) DNE |
||
| Issue Links: |
|
||||||||
| Severity: | 3 | ||||||||
| Rank (Obsolete): | 9223372036854775807 | ||||||||
| Description |
|
replay-single test_65b fails with 'No early reply'. We see this test fail frequently for PPC64 client testing. If we ignore the PPC failures, we see this test fail for DNE configurations from 16 FEB 2020 with https://testing.whamcloud.com/test_sets/9766a818-51c7-11ea-a90e-52540065bddc. In all of these cases, replay-single tests 0c and 0d fail with 'mount fails' and then a test after 65b will hang. Looking at the test_suite for a recent failure at https://testing.whamcloud.com/test_sets/2efe802a-db0c-4424-af23-e1a7e8ce99a4, we see the test output == replay-single test 65b: AT: verify early replies on packed reply / bulk =========================== 02:08:17 (1611540497) CMD: trevis-63vm4 lctl get_param -n at_max CMD: trevis-63vm1.trevis.whamcloud.com lctl get_param -n at_max CMD: trevis-63vm3 lctl get_param -n at_max CMD: trevis-63vm4 lctl get_param -n at_max CMD: trevis-63vm1.trevis.whamcloud.com lctl get_param -n at_max CMD: trevis-63vm3 lctl get_param -n at_max CMD: trevis-63vm4 lctl get_param -n at_history CMD: trevis-63vm4 lctl set_param at_history=8 at_history=8 CMD: trevis-63vm3 lctl set_param at_history=8 at_history=8 CMD: trevis-63vm4 /usr/sbin/lctl get_param -n debug debug=other trace CMD: trevis-63vm3 lctl set_param fail_val=6 fail_val=6 CMD: trevis-63vm3 /usr/sbin/lctl set_param fail_loc=0x224 fail_loc=0x224 CMD: trevis-63vm3 /usr/sbin/lctl set_param fail_loc=0 fail_loc=0 replay-single test_65b: @@@@@@ FAIL: No early reply Trace dump: = /usr/lib64/lustre/tests/test-framework.sh:6273:error() = /usr/lib64/lustre/tests/replay-single.sh:1864:test_65b() There are no errors in the console/dmesg logs. This issue maybe be related to LU-9566 replay-single test_65a: @@@@@@ FAIL: No early reply Logs for other failures are at |
| Comments |
| Comment by Andreas Dilger [ 07/Oct/22 ] |
|
Duplicate of LU-9566 |