Discussions io thread for glusterfs shd process

[mykaul]

@perrynzhou - please have the discussion here.

[perrynzhou]

@mohit84 @xhernandez @amarts @BarakSason @mohit84 @tshacked @csabahenk @pranithk we can discuss it

[pranithk]

I found when i execute gluster volume heal rep-vol full, how works for glusterfs shd ?is any document to describe that? i found > glusterfs shd collected a batch data that need to heal and do heal, is right?how to set debug log level for glusterfs shd?

heal full, crawls the filesystem and for each file/directory triggers heal.
Not sure about document. @itisravi Do you know?

I didn't find any way to change log-level of just shd. You can change log-level of both internal-clients(shd, rebalance) and external-clients(mounts) using

gluster volume set <volname> diagnostics.client-log-level DEBUG

[itisravi]

https://github.com/gluster/glusterfs/blob/devel/doc/developer-guide/afr-self-heal-daemon.md has some info. I wrote it long time ago, so it needs to be validated with current code. daemon-log-level was added to alter log level for shd and other daemons, so you can check if that works. See https://review.gluster.org/#/c/glusterfs/+/20442/