sebest / collectd-mongodb Goto Github PK

Instead of using a username/password authentication mechanism mongo supports key based authentication. This really should be supported as enterprises rely on key based auth in leu of old style username/pw auth.

Thank you for your consideration.

For available, current, and total connections, I get 0 unless I change connections to value:GAUGE:0:U. This is happening in collectd 5.4.1. Not sure how the behavior is in version 4.

Once I make the switch, I see the correct values reflected from db.serverStatus() now being stored in graphite.

Plugin is not compatible with pymongo 3.0 + as the only way to connect to mongo database is using MongoClient not by Connection(which is removed in 3.0 + ). And the link which is provided to install pymongo in documentation point to the master branch which is 3.0.1 version

Wouldn't it be a nice feature to let collectd know about all databases instead of a selection.

Maybe by a config entry.

I added mongodb.py to my collectd.conf file as shown here and it actually appears to be working but I am a bit concerned about this error I get when starting collectd. I did install mogodb.py in a different location than the one in the example, it's at /etc/collectd.d/mongodb.py. Here's the error:

[root@mongo6 collectd]# /etc/init.d/collectd restart
Stopping collectd: [ OK ]
Starting collectd: Parse error in file /etc/collectd.d/mongodb.py', line 1301 near': syntax error, unexpected EOL
yyparse returned error #1
configfile: Cannot read file /etc/collectd.d/mongodb.py'. Parse error in file/etc/collectd.d/mongodb.pyc', line 1318 near >': syntax error, unexpected CLOSEBRAC yyparse returned error #1 configfile: Cannot read file/etc/collectd.d/mongodb.pyc'.
.():():.=.=.=.=[,].=.=.=.=.=.=(,,,,=)::='%-%'%(.,):=(.)=.()[ OK ]=.=[,].()():=(=.,=.,=)=[.[]]..:.(.,.)=.('')=['']=()

So you can see that it starts and it does appear to be working as I have stats (that I don't quite trust because of the error) in graphite. Any ideas what's going on here?

system info:

[root@mongo6 collectd]# cat /etc/issue
Amazon Linux AMI release 2014.09
Kernel \r on an \m

[root@mongo6 collectd]# uname -a
Linux mongo6 3.10.42-52.145.amzn1.x86_64 #1 SMP Tue Jun 10 23:46:43 UTC 2014 x86_64 x86_64 x86_64 GNU/Linux

[root@mongo6 collectd]# rpm -q python
python-2.6-2.26.amzn1.noarch

[root@mongo6 collectd]# rpm -q collectd
collectd-5.4.1-1.5.amzn1.x86_64

[root@mongo6 collectd]# rpm -qa|grep -i mongo
mongodb-org-mongos-2.6.5-1.x86_64
mongodb-org-shell-2.6.5-1.x86_64
mongodb-org-tools-2.6.5-1.x86_64
mongodb-org-server-2.6.5-1.x86_64
mongodb-org-2.6.5-1.x86_64

The documentation says you can use the config option "Database", but this option isnt' available. The code suggests to use "Databases" but this doesn't work too:

Starting statistics collection and monitoring daemon: collectdUnhandled python exception in loading module: TypeError: cannot concatenate 'str' and 'tuple' objects

My config looks like this:

# This file autogenerated by Chef
# Do not edit, changes will be overwritten
<LoadPlugin python>
  Globals true
</LoadPlugin>

<Plugin python>
  ModulePath "/usr/lib/collectd"
  Import "mongodb"

  <Module "mongodb">
      Databases "admim" "mydb"
  </Module>
</Plugin>

I have fetched the latest code from this repo but I get an error message.

This is my config:

# This file autogenerated by Chef
# Do not edit, changes will be overwritten
<LoadPlugin python>
  Globals true
</LoadPlugin>

<Plugin python>
  ModulePath "/usr/lib/collectd"
  Import "mongodb"

  <Module "mongodb">

  </Module>
</Plugin>

MongoDB is working as you can see here:

root@db11:/var/log# mongo
MongoDB shell version: 2.4.6
connecting to: test
> db.stats()
{
        "db" : "test",
        "collections" : 0,
        "objects" : 0,
        "avgObjSize" : 0,
        "dataSize" : 0,
        "storageSize" : 0,
        "numExtents" : 0,
        "indexes" : 0,
        "indexSize" : 0,
        "fileSize" : 0,
        "nsSizeMB" : 0,
        "dataFileVersion" : {

        },
        "ok" : 1
}

Here are the error messages of collectd:

Jan  8 06:20:54 precise64 collectd[9176]: Unhandled python exception in init callback: TypeError: 'Database' object is not callable. If you meant to call the 'alive' method on a 'Connection' object it is failing because no such method exists.
Jan  8 06:20:54 precise64 collectd[9176]: Initialization complete, entering read-loop.
Jan  8 06:20:54 precise64 collectd[9176]: Unhandled python exception in read callback: TypeError: Dataset total_connections not found
Jan  8 06:20:54 precise64 collectd[9176]: read-function of plugin `python.mongodb' failed. Will suspend it for 20 seconds.
Jan  8 06:21:14 precise64 collectd[9176]: Unhandled python exception in read callback: TypeError: Dataset total_connections not found

I think I have configured the plugin correctly in collectd. However I'm not sure what to put into the /usr/share/collectd/types.db file, so that I won't get:

Unhandled python exception in read callback: TypeError: Dataset total_operations not found

in the logs at /var/log/messages

Any help would be much appreciated, thanks :)

I think I have everything configured correctly, but I had to add the following line to my config, or so I thought.

ModulePath "/usr/local/lib64/python2.6/site-packages/pymongo"

In /var/log/messages I see

python plugin: Error importing module "mongodb".
Unhandled python exception in importing module: ImportError: cannot import name Connection

I initially tried by just including the path to the plugin itself, but that was also failing with

Unhandled python exception in importing module: ImportError: No module named pymongo

Any ideas what I might be missing?

Hi,

We would like to evaluate your code to monitor MongoDB but are not sure what you code is licensed under.

Thanks,

Please update your plugin to support newest pymongo >= 4.3.3, because collectd issues different errors related to abandoned functionality.
To get stable work of the "collectd-mongodb" plugin I had to freeze pymongo == 3.13 version which seems to be obsolete.

Thanks !

This plugin opens a connection for every read. Would it make sense to leave it open? It seems like there'd be less overhead that way.

I've come across your plugin and am very interested in using it. But not knowing collectd and its plugins much I have no idea how to use it. Was wondering if you could perhaps include a small section to describe the installation/usage process?

Thanks :)

I just wanted to let you know that I've confirmed that this works with Collectd 5.2.0 and MongoDB 2.4.5. You may want to update the README.

Thanks!

Hi,

Added this plugin and the data seems to be coming in for this, however noticed on the connection value its bringing back 00.1 as the figure, not sure how you can have 00.5 of a connection I would have thought this would be a total any ideas?

Details:
[{"target": "mongodb-01.livemongo001_live_dc1_server.mongo.27017.connections.connections.value", "datapoints": [[0.0, 1398867120], [0.0, 1398867180], [0.0, 1398867240], [0.0, 1398867300], [null, 1398867360]]}, {"target": "mongodb-01.livemongo002_live_dc1_server.mongo.27017.connections.connections.value", "datapoints": [[0.0, 1398867120], [0.0, 1398867180], [0.016664999999999999, 1398867240], [0.0, 1398867300], [null, 1398867360]]}, {"target": "mongodb-01.livemongo003_live_dc1_server.mongo.27017.connections.connections.value", "datapoints": [[0.0, 1398867120], [0.016667999999999999, 1398867180], [0.016664000000000002, 1398867240], [0.0, 1398867300], [null, 1398867360]]}, {"target": "mongodb-01.livemongo004_live_dc2_server.mongo.27017.connections.connections.value", "datapoints": [[0.0, 1398867120], [0.016667000000000001, 1398867180], [0.0, 1398867240], [0.0, 1398867300], [0.0, 1398867360]]}, {"target": "mongodb-01.livemongo005_live_dc2_server.mongo.27017.connections.connections.value", "datapoints": [[0.0, 1398867120], [0.0, 1398867180], [0.0, 1398867240], [0.0, 1398867300], [0.0, 1398867360]]}]