From: John Spray <john.spray@redhat.com>
Date: Wed, 20 May 2015 15:40:04 +0000 (+0100)
Subject: mds: fix MDLog error handling on ENOENT
X-Git-Tag: v9.0.3~132^2~7
X-Git-Url: http://git-server-git.apps.pok.os.sepia.ceph.com/?a=commitdiff_plain;h=1165373e8869d2b78c5aef2c1bec9f8869c4e85a;p=ceph.git

mds: fix MDLog error handling on ENOENT

This was trying to check for standby replay
by testing is_readonly(), but that's always
true during normal replay.  Check mds->is_standby_replay()
instead.

Signed-off-by: John Spray <john.spray@redhat.com>
---

diff --git a/src/mds/MDLog.cc b/src/mds/MDLog.cc
index 4c22e7e4a92a..4ee502ee1a80 100644
--- a/src/mds/MDLog.cc
+++ b/src/mds/MDLog.cc
@@ -1145,7 +1145,7 @@ void MDLog::_replay_thread()
       r = journaler->get_error();
       dout(0) << "_replay journaler got error " << r << ", aborting" << dendl;
       if (r == -ENOENT) {
-        if (journaler->is_readonly()) {
+        if (mds->is_standby_replay()) {
           // journal has been trimmed by somebody else
           r = -EAGAIN;
         } else {