[Pkg-nagios-changes] [pkg-nagios-plugins-contrib] 01/04: check_mongo: Update to recent version
Jan Wagner
waja at moszumanska.debian.org
Tue Jun 23 12:31:49 UTC 2015
This is an automated email from the git hooks/post-receive script.
waja pushed a commit to branch master
in repository pkg-nagios-plugins-contrib.
commit c11ef666bfa574a1d9abb57a04575ed0d8cda1d1
Author: Jan Wagner <waja at cyconet.org>
Date: Tue Jun 23 12:50:30 2015 +0200
check_mongo: Update to recent version
---
check_mongodb/check_mongodb.py | 123 ++++++++++++++++++++++++++---------------
check_mongodb/control | 2 +-
2 files changed, 80 insertions(+), 45 deletions(-)
diff --git a/check_mongodb/check_mongodb.py b/check_mongodb/check_mongodb.py
index 87b4ce0..1586f0b 100644
--- a/check_mongodb/check_mongodb.py
+++ b/check_mongodb/check_mongodb.py
@@ -17,7 +17,7 @@
# - Dag Stockstad <dag.stockstad at gmail.com>
# - @Andor on github
# - Steven Richards - Captainkrtek on github
-# - Max Vernimmen
+# - Max Vernimmen - @mvernimmen-CG / @mvernimmen on github
#
# USAGE
#
@@ -145,6 +145,8 @@ def main(argv):
p.add_option('-q', '--querytype', action='store', dest='query_type', default='query', help='The query type to check [query|insert|update|delete|getmore|command] from queries_per_second')
p.add_option('-c', '--collection', action='store', dest='collection', default='admin', help='Specify the collection to check')
p.add_option('-T', '--time', action='store', type='int', dest='sample_time', default=1, help='Time used to sample number of pages faults')
+ p.add_option('-M', '--mongoversion', action='store', type='choice', dest='mongo_version', default='2', help='The MongoDB version you are talking with, either 2 or 3',
+ choices=['2','3'])
options, arguments = p.parse_args()
host = options.host
@@ -164,6 +166,7 @@ def main(argv):
action = options.action
perf_data = options.perf_data
max_lag = options.max_lag
+ mongo_version = options.mongo_version
database = options.database
ssl = options.ssl
replicaset = options.replicaset
@@ -193,13 +196,13 @@ def main(argv):
elif action == "replset_state":
return check_replset_state(con, perf_data, warning, critical)
elif action == "memory":
- return check_memory(con, warning, critical, perf_data, options.mapped_memory)
+ return check_memory(con, warning, critical, perf_data, options.mapped_memory, host)
elif action == "memory_mapped":
return check_memory_mapped(con, warning, critical, perf_data)
elif action == "queues":
return check_queues(con, warning, critical, perf_data)
elif action == "lock":
- return check_lock(con, warning, critical, perf_data)
+ return check_lock(con, warning, critical, perf_data, mongo_version)
elif action == "current_lock":
return check_current_lock(con, host, warning, critical, perf_data)
elif action == "flushing":
@@ -236,9 +239,9 @@ def main(argv):
elif action == "asserts":
return check_asserts(con, host, warning, critical, perf_data)
elif action == "replica_primary":
- return check_replica_primary(con, host, warning, critical, perf_data, replicaset)
+ return check_replica_primary(con, host, warning, critical, perf_data, replicaset, mongo_version)
elif action == "queries_per_second":
- return check_queries_per_second(con, query_type, warning, critical, perf_data)
+ return check_queries_per_second(con, query_type, warning, critical, perf_data, mongo_version)
elif action == "page_faults":
check_page_faults(con, sample_time, warning, critical, perf_data)
elif action == "chunks_balance":
@@ -337,6 +340,10 @@ def check_connections(con, warning, critical, perf_data):
def check_rep_lag(con, host, port, warning, critical, percent, perf_data, max_lag, user, passwd):
# Get mongo to tell us replica set member name when connecting locally
if "127.0.0.1" == host:
+ if not "me" in con.admin.command("ismaster","1").keys():
+ print "OK - This is not replicated MongoDB"
+ sys.exit(3)
+
host = con.admin.command("ismaster","1")["me"].split(':')[0]
if percent:
@@ -354,7 +361,7 @@ def check_rep_lag(con, host, port, warning, critical, percent, perf_data, max_la
try:
rs_status = con.admin.command("replSetGetStatus")
except pymongo.errors.OperationFailure, e:
- if e.code == None and str(e).find('failed: not running with --replSet"'):
+ if ((e.code == None and str(e).find('failed: not running with --replSet"')) or (e.code == 76 and str(e).find('not running with --replSet"'))):
print "OK - Not running with replSet"
return 0
@@ -498,23 +505,31 @@ def check_rep_lag(con, host, port, warning, critical, percent, perf_data, max_la
except Exception, e:
return exit_with_general_critical(e)
-
-def check_memory(con, warning, critical, perf_data, mapped_memory):
- #
- # These thresholds are basically meaningless, and must be customized to your system's ram
- #
-
- # Get the total system merory and calculate based on that how much memory used by Mongodb is ok or not.
+#
+# Check the memory usage of mongo. Alerting on this may be hard to get right
+# because it'll try to get as much memory as it can. And that's probably
+# a good thing.
+#
+def check_memory(con, warning, critical, perf_data, mapped_memory, host):
+ # Get the total system memory of this system (This is totally bogus if you
+ # are running this command remotely) and calculate based on that how much
+ # memory used by Mongodb is ok or not.
meminfo = open('/proc/meminfo').read()
matched = re.search(r'^MemTotal:\s+(\d+)', meminfo)
if matched:
mem_total_kB = int(matched.groups()[0])
- # Old way
- #critical = critical or 16
- # The new way. if using >80% then warn, if >90% then critical level
- warning = warning or (mem_total_kB * 0.8) / 1024.0 / 1024.0
- critical = critical or (mem_total_kB * 0.9) / 1024.0 / 1024.0
+ if host != "127.0.0.1" and not warning:
+ # Running remotely and value was not set by user, use hardcoded value
+ warning = 12
+ else:
+ # running locally or user provided value
+ warning = warning or (mem_total_kB * 0.8) / 1024.0 / 1024.0
+
+ if host != "127.0.0.1" and not critical:
+ critical = 16
+ else:
+ critical = critical or (mem_total_kB * 0.9) / 1024.0 / 1024.0
# debugging
#print "mem total: {0}kb, warn: {1}GB, crit: {2}GB".format(mem_total_kB,warning, critical)
@@ -589,7 +604,7 @@ def check_memory_mapped(con, warning, critical, perf_data):
message += " %.2fGB mappedWithJournal" % mem_mapped_journal
except:
mem_mapped_journal = 0
- message += performance_data(perf_data, [("%.2f" % mem_mapped, "memory_mapped"), ("%.2f" % mem_mapped_journal, "mappedWithJournal")])
+ message += performance_data(perf_data, [("%.2f" % mem_mapped, "memory_mapped", warning, critical), ("%.2f" % mem_mapped_journal, "mappedWithJournal")])
if not mem_mapped == -1:
return check_levels(mem_mapped, warning, critical, message)
@@ -601,26 +616,33 @@ def check_memory_mapped(con, warning, critical, perf_data):
return exit_with_general_critical(e)
-def check_lock(con, warning, critical, perf_data):
+#
+# Return the percentage of the time there was a global Lock
+#
+def check_lock(con, warning, critical, perf_data, mongo_version):
warning = warning or 10
critical = critical or 30
- try:
- data = get_server_status(con)
- #
- # calculate percentage
- #
- lockTime = data['globalLock']['lockTime']
- totalTime = data['globalLock']['totalTime']
- if lockTime > totalTime:
- lock_percentage = 0.00
- else:
- lock_percentage = float(lockTime) / float(totalTime) * 100
- message = "Lock Percentage: %.2f%%" % lock_percentage
- message += performance_data(perf_data, [("%.2f" % lock_percentage, "lock_percentage", warning, critical)])
- return check_levels(lock_percentage, warning, critical, message)
-
- except Exception, e:
- return exit_with_general_critical(e)
+ if mongo_version == "2":
+ try:
+ data = get_server_status(con)
+ lockTime = data['globalLock']['lockTime']
+ totalTime = data['globalLock']['totalTime']
+ #
+ # calculate percentage
+ #
+ if lockTime > totalTime:
+ lock_percentage = 0.00
+ else:
+ lock_percentage = float(lockTime) / float(totalTime) * 100
+ message = "Lock Percentage: %.2f%%" % lock_percentage
+ message += performance_data(perf_data, [("%.2f" % lock_percentage, "lock_percentage", warning, critical)])
+ return check_levels(lock_percentage, warning, critical, message)
+ except Exception, e:
+ print "Couldn't get globalLock lockTime info from mongo, are you sure you're not using version 3? See the -M option."
+ return exit_with_general_critical(e)
+ else:
+ print "FAIL - Mongo3 doesn't report on global locks"
+ return 1
def check_flushing(con, warning, critical, avg, perf_data):
@@ -723,7 +745,7 @@ def check_replset_state(con, perf_data, warning="", critical=""):
data = con.admin.command(son.SON([('replSetGetStatus', 1)]))
state = int(data['myState'])
except pymongo.errors.OperationFailure, e:
- if e.code == None and str(e).find('failed: not running with --replSet"'):
+ if ((e.code == None and str(e).find('failed: not running with --replSet"')) or (e.code == 76 and str(e).find('not running with --replSet"'))):
state = -1
if state == 8:
@@ -934,7 +956,7 @@ def check_collection_size(con, database, collection, warning, critical, perf_dat
except Exception, e:
return exit_with_general_critical(e)
-def check_queries_per_second(con, query_type, warning, critical, perf_data):
+def check_queries_per_second(con, query_type, warning, critical, perf_data, mongo_version):
warning = warning or 250
critical = critical or 500
@@ -958,7 +980,10 @@ def check_queries_per_second(con, query_type, warning, critical, perf_data):
query_per_sec = float(diff_query) / float(diff_ts)
# update the count now
- db.nagios_check.update({u'_id': last_count['_id']}, {'$set': {"data.%s" % query_type: {'count': num, 'ts': int(time.time())}}})
+ if mongo_version == "2":
+ db.nagios_check.update({u'_id': last_count['_id']}, {'$set': {"data.%s" % query_type: {'count': num, 'ts': int(time.time())}}})
+ else:
+ db.nagios_check.update_one({u'_id': last_count['_id']}, {'$set': {"data.%s" % query_type: {'count': num, 'ts': int(time.time())}}})
message = "Queries / Sec: %f" % query_per_sec
message += performance_data(perf_data, [(query_per_sec, "%s_per_sec" % query_type, warning, critical, message)])
@@ -967,13 +992,20 @@ def check_queries_per_second(con, query_type, warning, critical, perf_data):
# since it is the first run insert it
query_per_sec = 0
message = "First run of check.. no data"
- db.nagios_check.update({u'_id': last_count['_id']}, {'$set': {"data.%s" % query_type: {'count': num, 'ts': int(time.time())}}})
+ if mongo_version == "2":
+ db.nagios_check.update({u'_id': last_count['_id']}, {'$set': {"data.%s" % query_type: {'count': num, 'ts': int(time.time())}}})
+ else:
+ db.nagios_check.update_one({u'_id': last_count['_id']}, {'$set': {"data.%s" % query_type: {'count': num, 'ts': int(time.time())}}})
+
except TypeError:
#
# since it is the first run insert it
query_per_sec = 0
message = "First run of check.. no data"
- db.nagios_check.insert({'check': 'query_counts', 'data': {query_type: {'count': num, 'ts': int(time.time())}}})
+ if mongo_version == "2":
+ db.nagios_check.insert({'check': 'query_counts', 'data': {query_type: {'count': num, 'ts': int(time.time())}}})
+ else:
+ db.nagios_check.insert_one({'check': 'query_counts', 'data': {query_type: {'count': num, 'ts': int(time.time())}}})
return check_levels(query_per_sec, warning, critical, message)
@@ -1206,7 +1238,7 @@ def get_stored_primary_server_name(db):
return stored_primary_server
-def check_replica_primary(con, host, warning, critical, perf_data, replicaset):
+def check_replica_primary(con, host, warning, critical, perf_data, replicaset, mongo_version):
""" A function to check if the primary server of a replica set has changed """
if warning is None and critical is None:
warning = 1
@@ -1229,7 +1261,10 @@ def check_replica_primary(con, host, warning, critical, perf_data, replicaset):
saved_primary = "None"
if current_primary != saved_primary:
last_primary_server_record = {"server": current_primary}
- db.last_primary_server.update({"_id": "last_primary"}, {"$set": last_primary_server_record}, upsert=True, safe=True)
+ if mongo_version == "2":
+ db.last_primary_server.update({"_id": "last_primary"}, {"$set": last_primary_server_record}, upsert=True, safe=True)
+ else:
+ db.last_primary_server.update_one({"_id": "last_primary"}, {"$set": last_primary_server_record}, upsert=True, safe=True)
message = "Primary server has changed from %s to %s" % (saved_primary, current_primary)
primary_status = 1
return check_levels(primary_status, warning, critical, message)
diff --git a/check_mongodb/control b/check_mongodb/control
index 4a69ff8..f0508d3 100644
--- a/check_mongodb/control
+++ b/check_mongodb/control
@@ -1,6 +1,6 @@
Uploaders: Jan Wagner <waja at cyconet.org>
Recommends: python-pymongo
-Version: 2032d64ba1
+Version: 19e6b2c8f4
Homepage: https://github.com/mzupan/nagios-plugin-mongodb
Watch: https://github.com/mzupan/nagios-plugin-mongodb <span class="sha">([0-9a-f]+)</span>
Description: Plugin script to monitor your MongoDB server(s)
--
Alioth's /usr/local/bin/git-commit-notice on /srv/git.debian.org/git/pkg-nagios/pkg-nagios-plugins-contrib.git
More information about the Pkg-nagios-changes
mailing list