[Bugs] [Bug 1287517] Memory leak in glusterd

[bugzilla at redhat.com]

https://bugzilla.redhat.com/show_bug.cgi?id=1287517



--- Comment #41 from Roman Tereshonkov <tereshonkov.roman at gmail.com> ---
It looks like it is a functional problem.

The issued command "gluster volume status all" generates a series of
transaction identified by different ids.
The very first one is to get the volume names supported by remote host
glusterfs server.
The others transaction are issued per each volume name and include
locking/unlocking phases.
And it is the unlocking phase that is the marker that the transaction ended and
all allocated resources are to be free.
There is no other code pathway to call glusterd_clear_txn_opinfo function
except unlock or unlock_all pathways.
Thus, there is no any way for the very first transaction to know when it is
ended to call glusterd_clear_txn_opinfo and free opinfo.

This is my analysis which is needed to be confirmed by you.
Logically it is different issue now from what we have already fixed.

Should I open new bug for this functional problem?

Do you see any simple way to fix it?

-- 
You are receiving this mail because:
You are on the CC list for the bug.
Unsubscribe from this bug https://bugzilla.redhat.com/token.cgi?t=f0CXl7WBLK&a=cc_unsubscribe