From b7309a0d536875b7e79493dbe09b8e34cb3e387a Mon Sep 17 00:00:00 2001
From: Iustin Pop <iustin@google.com>
Date: Wed, 1 Oct 2008 09:27:17 +0000
Subject: [PATCH] Fix the watcher not restarting instance bug
The watcher was using conflicting attributes of the instance:
- it queried the admin_/oper_state, which are booleans
- but it compared those to the status (which is a text field)
The code was changed to query the aggregated 'status' field, as that
will also return indication of node problems, and we can use this only
one field for all decisions. We still ask for the admin_state field as
that is needed for the activate disks check (in secondary node restart).
The patch also touches the watcher in some other parts:
- log exceptions nicer
- convert a method to @staticmethod
- remove unused imports
Reviewed-by: imsnah
---
daemons/ganeti-watcher | 24 ++++++++++--------------
1 file changed, 10 insertions(+), 14 deletions(-)
diff --git a/daemons/ganeti-watcher b/daemons/ganeti-watcher
index 94c6b48bb..438cf00ad 100755
--- a/daemons/ganeti-watcher
+++ b/daemons/ganeti-watcher
@@ -30,8 +30,6 @@ by a node reboot. Run from cron or similar.
import os
import sys
import time
-import fcntl
-import errno
import logging
from optparse import OptionParser
@@ -46,8 +44,8 @@ from ganeti import cli
MAXTRIES = 5
-BAD_STATES = ['stopped']
-HELPLESS_STATES = ['(node down)']
+BAD_STATES = ['ERROR_down']
+HELPLESS_STATES = ['ERROR_nodedown']
NOTICE = 'NOTICE'
ERROR = 'ERROR'
KEY_RESTART_COUNT = "restart_count"
@@ -238,7 +236,7 @@ def GetInstanceList(with_secondaries=None):
"""Get a list of instances on this cluster.
"""
- fields = ["name", "oper_state", "admin_state"]
+ fields = ["name", "status", "admin_state"]
if with_secondaries is not None:
fields.append("snodes")
@@ -327,8 +325,9 @@ class Watcher(object):
try:
logging.info("Activating disks for instance %s", instance.name)
instance.ActivateDisks()
- except Exception, err:
- logging.error(str(err), exc_info=True)
+ except Exception:
+ logging.exception("Error while activating disks for instance %s",
+ instance.name)
# Keep changed boot IDs
for name in check_nodes:
@@ -339,10 +338,6 @@ class Watcher(object):
"""
for instance in self.instances:
- # Don't care about manually stopped instances
- if not instance.autostart:
- continue
-
if instance.state in BAD_STATES:
n = notepad.NumberOfRestartAttempts(instance)
@@ -361,8 +356,8 @@ class Watcher(object):
instance.name, last)
instance.Restart()
self.started_instances.add(instance.name)
- except Exception, err:
- logging.error(str(err), exc_info=True)
+ except Exception:
+ logging.exception("Erro while restarting instance %s", instance.name)
notepad.RecordRestartAttempt(instance)
elif instance.state in HELPLESS_STATES:
@@ -373,7 +368,8 @@ class Watcher(object):
notepad.RemoveInstance(instance)
logging.info("Restart of %s succeeded", instance.name)
- def VerifyDisks(self):
+ @staticmethod
+ def VerifyDisks():
"""Run gnt-cluster verify-disks.
"""
--
GitLab