Conversation
|
Ticket title is 'Infinite DER_NOTLEADER loop on single-server DAOS cluster after pool create' |
Signed-off-by: Li Wei <liwei@hpe.com>
|
Test stage NLT completed with status FAILURE. https://jenkins-3.daos.hpc.amslabs.hpecorp.net/job/daos-stack/job/daos/job/PR-18016/1/display/redirect |
|
Test stage Unit Test completed with status FAILURE. https://jenkins-3.daos.hpc.amslabs.hpecorp.net/job/daos-stack/job/daos/job/PR-18016/1/display/redirect |
|
Mark as ready for reviews, for the CI testing is too slow at the moment. |
|
Test stage Functional Hardware Medium MD on SSD completed with status UNSTABLE. https://jenkins-3.daos.hpc.amslabs.hpecorp.net/job/daos-stack/job/daos//view/change-requests/job/PR-18016/2/testReport/ |
|
Build 2:
|
wangshilong
left a comment
There was a problem hiding this comment.
A bit tricky, but looks workable. So if ->iv_valid is true it means IV is updated. but It could be empty/dummy value.
|
Probably should go 2.8 as well, NOT leader errors are confusing and annoying |
It appears that if the PS leader skips the ds_pool_iv_conn_hdls_update
call during pool_svc_step_up_cb because there's no pool handle in the
DB, IV fetches for pool handles will create invalid IV entries and
return unexpected -DER_NOTLEADERs. To prevent that, this patch changes
pool_svc_step_up_cb to call ds_pool_iv_conn_hdls_update even if there's
no pool handle in the DB.
Steps for the author:
After all prior steps are complete: