Force user to grow metadata, if user manually extends datalv of a thinpool and the current metadata size will not be enough to address the new size pool
Descriptionnikhil kshirsagar
2019-02-12 04:03:03 UTC
Description of problem:
This bz created out of similar RHEL7 bz https://bugzilla.redhat.com/show_bug.cgi?id=1676359
As discussed on https://docs.google.com/document/d/15J2tSBc3scl_Gtl4dbhFUe6ApPo90yrlAJZ6e59i4z0 point 2, we must force (or at least strongly warn) users to grow metadata manually too, if they manually grow the thinpool data using lvextend. This will prevent metadata full issues in the future.
Additional info:
Many gluster customers have small pools created by default scripts during setup, and have autoextend turned off by design, because gluster wants to be in control of the storage allocation, and doesn't allow autoextension because that would circumvent their storage manager. This results in customers manually increasing data subsequently, but many times forgetting to increase metadata too. Eventually metadata reaches 100% and thinpool corruptions ensue.
Comment 8RHEL Program Management
2021-02-01 07:32:41 UTC
After evaluating this issue, there are no plans to address it further or fix it in an upcoming release. Therefore, it is being closed. If plans change such that this issue will be fixed in an upcoming release, then the bug can be reopened.