Experience with monitoring cookbooks?

Ohai, Masterchefs!

We’re in the process of standing up a Chef installation, using it to provision and manage our servers and a variety of packages, and I’ve gotten down to the point where I’m now working on the monitoring component. I know there are lots of monitoring cookbooks on the opscode site, and I’m wondering if you folks can share with me your experience in using the various cookbooks in this area and what you like or don’t like about whatever packages you’ve tried.

Our particular environment is CentOS 5.6 on Rackspace cloud servers at the moment, but I’d love to hear experiences regardless of OS and type of installation.

I did notice that Zenoss recently came out with a new zenpack so that they can monitor your OpenStack systems, including things like how many flavors you have available and how many of each type are in use, how many images are available and how many are in use, what the various server states are, etc… This is in addition to what they can monitor at the individual server level, if you’re also including that as well.

Has anyone played with this new zenpack and can give me some feedback about how it has impacted your site?

Thanks!


Brad Knowles bknowles@ihiji.com
SAGE Level IV, Chef Level 0.0.1

I'd suggest to try them.

The munin cookbook works great but I had terrible issues with munin on
CentOS Rackspace (one core blocked at 100%, need to reboot).

I found Zenoss is extremely confusing and difficult to navigate.

We're in the process of setting up ganglia/graphite with Chef on our
system soon.

Anyway try them and pick what you think is right for you.

--Gilles

On Wed, Sep 21, 2011 at 9:22 AM, Brad Knowles bknowles@ihiji.com wrote:

Ohai, Masterchefs!

We're in the process of standing up a Chef installation, using it to provision and manage our servers and a variety of packages, and I've gotten down to the point where I'm now working on the monitoring component. I know there are lots of monitoring cookbooks on the opscode site, and I'm wondering if you folks can share with me your experience in using the various cookbooks in this area and what you like or don't like about whatever packages you've tried.

Our particular environment is CentOS 5.6 on Rackspace cloud servers at the moment, but I'd love to hear experiences regardless of OS and type of installation.

I did notice that Zenoss recently came out with a new zenpack so that they can monitor your OpenStack systems, including things like how many flavors you have available and how many of each type are in use, how many images are available and how many are in use, what the various server states are, etc.... This is in addition to what they can monitor at the individual server level, if you're also including that as well.

Has anyone played with this new zenpack and can give me some feedback about how it has impacted your site?

Thanks!

--
Brad Knowles bknowles@ihiji.com
SAGE Level IV, Chef Level 0.0.1

Anyone looked into using opentsdb or have a chef recipe for it? I
like the approach of not using summarization. Lots of cheap storage
scales better than clever solutions in a high growth situation (in my
experience anyway).

-Peter

On Thu, Sep 22, 2011 at 12:16 PM, Gilles Devaux gilles.devaux@gmail.com wrote:

I'd suggest to try them.

The munin cookbook works great but I had terrible issues with munin on
CentOS Rackspace (one core blocked at 100%, need to reboot).

I found Zenoss is extremely confusing and difficult to navigate.

We're in the process of setting up ganglia/graphite with Chef on our
system soon.

Anyway try them and pick what you think is right for you.

--Gilles

On Wed, Sep 21, 2011 at 9:22 AM, Brad Knowles bknowles@ihiji.com wrote:

Ohai, Masterchefs!

We're in the process of standing up a Chef installation, using it to provision and manage our servers and a variety of packages, and I've gotten down to the point where I'm now working on the monitoring component. I know there are lots of monitoring cookbooks on the opscode site, and I'm wondering if you folks can share with me your experience in using the various cookbooks in this area and what you like or don't like about whatever packages you've tried.

Our particular environment is CentOS 5.6 on Rackspace cloud servers at the moment, but I'd love to hear experiences regardless of OS and type of installation.

I did notice that Zenoss recently came out with a new zenpack so that they can monitor your OpenStack systems, including things like how many flavors you have available and how many of each type are in use, how many images are available and how many are in use, what the various server states are, etc.... This is in addition to what they can monitor at the individual server level, if you're also including that as well.

Has anyone played with this new zenpack and can give me some feedback about how it has impacted your site?

Thanks!

--
Brad Knowles bknowles@ihiji.com
SAGE Level IV, Chef Level 0.0.1

We looked at opentsdb and the problem is that we can't find any
tcollector besides the one here: TCollector - OpenTSDB - A Distributed, Scalable Monitoring System

So if you want to monitor haproxy or mongo / riak, etc... you'll have
to write them. You can find ganglia gmetric for pretty much
everything.

On Thu, Sep 22, 2011 at 9:26 AM, Peter Norton pn+chef-list@knewton.com wrote:

Anyone looked into using opentsdb or have a chef recipe for it? I
like the approach of not using summarization. Lots of cheap storage
scales better than clever solutions in a high growth situation (in my
experience anyway).

-Peter

On Thu, Sep 22, 2011 at 12:16 PM, Gilles Devaux gilles.devaux@gmail.com wrote:

I'd suggest to try them.

The munin cookbook works great but I had terrible issues with munin on
CentOS Rackspace (one core blocked at 100%, need to reboot).

I found Zenoss is extremely confusing and difficult to navigate.

We're in the process of setting up ganglia/graphite with Chef on our
system soon.

Anyway try them and pick what you think is right for you.

--Gilles

On Wed, Sep 21, 2011 at 9:22 AM, Brad Knowles bknowles@ihiji.com wrote:

Ohai, Masterchefs!

We're in the process of standing up a Chef installation, using it to provision and manage our servers and a variety of packages, and I've gotten down to the point where I'm now working on the monitoring component. I know there are lots of monitoring cookbooks on the opscode site, and I'm wondering if you folks can share with me your experience in using the various cookbooks in this area and what you like or don't like about whatever packages you've tried.

Our particular environment is CentOS 5.6 on Rackspace cloud servers at the moment, but I'd love to hear experiences regardless of OS and type of installation.

I did notice that Zenoss recently came out with a new zenpack so that they can monitor your OpenStack systems, including things like how many flavors you have available and how many of each type are in use, how many images are available and how many are in use, what the various server states are, etc.... This is in addition to what they can monitor at the individual server level, if you're also including that as well.

Has anyone played with this new zenpack and can give me some feedback about how it has impacted your site?

Thanks!

--
Brad Knowles bknowles@ihiji.com
SAGE Level IV, Chef Level 0.0.1

I use the Nagios cookbook and Ganglia cookbook and I'm very impressed with
both (I have hacked both quite a bit though)

On Thu, Sep 22, 2011 at 5:40 PM, Gilles Devaux gilles.devaux@gmail.comwrote:

We looked at opentsdb and the problem is that we can't find any
tcollector besides the one here: TCollector - OpenTSDB - A Distributed, Scalable Monitoring System

So if you want to monitor haproxy or mongo / riak, etc... you'll have
to write them. You can find ganglia gmetric for pretty much
everything.

On Thu, Sep 22, 2011 at 9:26 AM, Peter Norton pn+chef-list@knewton.com
wrote:

Anyone looked into using opentsdb or have a chef recipe for it? I
like the approach of not using summarization. Lots of cheap storage
scales better than clever solutions in a high growth situation (in my
experience anyway).

-Peter

On Thu, Sep 22, 2011 at 12:16 PM, Gilles Devaux gilles.devaux@gmail.com
wrote:

I'd suggest to try them.

The munin cookbook works great but I had terrible issues with munin on
CentOS Rackspace (one core blocked at 100%, need to reboot).

I found Zenoss is extremely confusing and difficult to navigate.

We're in the process of setting up ganglia/graphite with Chef on our
system soon.

Anyway try them and pick what you think is right for you.

--Gilles

On Wed, Sep 21, 2011 at 9:22 AM, Brad Knowles bknowles@ihiji.com
wrote:

Ohai, Masterchefs!

We're in the process of standing up a Chef installation, using it to
provision and manage our servers and a variety of packages, and I've gotten
down to the point where I'm now working on the monitoring component. I know
there are lots of monitoring cookbooks on the opscode site, and I'm
wondering if you folks can share with me your experience in using the
various cookbooks in this area and what you like or don't like about
whatever packages you've tried.

Our particular environment is CentOS 5.6 on Rackspace cloud servers at
the moment, but I'd love to hear experiences regardless of OS and type of
installation.

I did notice that Zenoss recently came out with a new zenpack so that
they can monitor your OpenStack systems, including things like how many
flavors you have available and how many of each type are in use, how many
images are available and how many are in use, what the various server states
are, etc.... This is in addition to what they can monitor at the individual
server level, if you're also including that as well.

Has anyone played with this new zenpack and can give me some feedback
about how it has impacted your site?

Thanks!

--
Brad Knowles bknowles@ihiji.com
SAGE Level IV, Chef Level 0.0.1

Ah, so the community isn't there yet, it's a momentum issue with not
having enough packaged collectors?

Thanks,

-Peter

On Thu, Sep 22, 2011 at 12:40 PM, Gilles Devaux gilles.devaux@gmail.com wrote:

We looked at opentsdb and the problem is that we can't find any
tcollector besides the one here: TCollector - OpenTSDB - A Distributed, Scalable Monitoring System

So if you want to monitor haproxy or mongo / riak, etc... you'll have
to write them. You can find ganglia gmetric for pretty much
everything.

On Thu, Sep 22, 2011 at 9:26 AM, Peter Norton pn+chef-list@knewton.com wrote:

Anyone looked into using opentsdb or have a chef recipe for it? I
like the approach of not using summarization. Lots of cheap storage
scales better than clever solutions in a high growth situation (in my
experience anyway).

-Peter

On Thu, Sep 22, 2011 at 12:16 PM, Gilles Devaux gilles.devaux@gmail.com wrote:

I'd suggest to try them.

The munin cookbook works great but I had terrible issues with munin on
CentOS Rackspace (one core blocked at 100%, need to reboot).

I found Zenoss is extremely confusing and difficult to navigate.

We're in the process of setting up ganglia/graphite with Chef on our
system soon.

Anyway try them and pick what you think is right for you.

--Gilles

On Wed, Sep 21, 2011 at 9:22 AM, Brad Knowles bknowles@ihiji.com wrote:

Ohai, Masterchefs!

We're in the process of standing up a Chef installation, using it to provision and manage our servers and a variety of packages, and I've gotten down to the point where I'm now working on the monitoring component. I know there are lots of monitoring cookbooks on the opscode site, and I'm wondering if you folks can share with me your experience in using the various cookbooks in this area and what you like or don't like about whatever packages you've tried.

Our particular environment is CentOS 5.6 on Rackspace cloud servers at the moment, but I'd love to hear experiences regardless of OS and type of installation.

I did notice that Zenoss recently came out with a new zenpack so that they can monitor your OpenStack systems, including things like how many flavors you have available and how many of each type are in use, how many images are available and how many are in use, what the various server states are, etc.... This is in addition to what they can monitor at the individual server level, if you're also including that as well.

Has anyone played with this new zenpack and can give me some feedback about how it has impacted your site?

Thanks!

--
Brad Knowles bknowles@ihiji.com
SAGE Level IV, Chef Level 0.0.1

We use the nagios cookbook and it works fairly well, have had to make
a bunch of adjustments to make it work well in our environment though.
I recommend the nagios quickstart guide.

I've tried the ganglia cookbook but haven't had much success getting
everything to "just work".

The munin cookbook works out of the box, but munin doesn't seem to
work very well with anything more than a small infrastructure.

-Mike

On Thu, Sep 22, 2011 at 9:51 AM, william pink will.pink@gmail.com wrote:

I use the Nagios cookbook and Ganglia cookbook and I'm very impressed with
both (I have hacked both quite a bit though)

On Thu, Sep 22, 2011 at 5:40 PM, Gilles Devaux gilles.devaux@gmail.com
wrote:

We looked at opentsdb and the problem is that we can't find any
tcollector besides the one here: TCollector - OpenTSDB - A Distributed, Scalable Monitoring System

So if you want to monitor haproxy or mongo / riak, etc... you'll have
to write them. You can find ganglia gmetric for pretty much
everything.

On Thu, Sep 22, 2011 at 9:26 AM, Peter Norton pn+chef-list@knewton.com
wrote:

Anyone looked into using opentsdb or have a chef recipe for it? I
like the approach of not using summarization. Lots of cheap storage
scales better than clever solutions in a high growth situation (in my
experience anyway).

-Peter

On Thu, Sep 22, 2011 at 12:16 PM, Gilles Devaux
gilles.devaux@gmail.com wrote:

I'd suggest to try them.

The munin cookbook works great but I had terrible issues with munin on
CentOS Rackspace (one core blocked at 100%, need to reboot).

I found Zenoss is extremely confusing and difficult to navigate.

We're in the process of setting up ganglia/graphite with Chef on our
system soon.

Anyway try them and pick what you think is right for you.

--Gilles

On Wed, Sep 21, 2011 at 9:22 AM, Brad Knowles bknowles@ihiji.com
wrote:

Ohai, Masterchefs!

We're in the process of standing up a Chef installation, using it to
provision and manage our servers and a variety of packages, and I've gotten
down to the point where I'm now working on the monitoring component. I know
there are lots of monitoring cookbooks on the opscode site, and I'm
wondering if you folks can share with me your experience in using the
various cookbooks in this area and what you like or don't like about
whatever packages you've tried.

Our particular environment is CentOS 5.6 on Rackspace cloud servers at
the moment, but I'd love to hear experiences regardless of OS and type of
installation.

I did notice that Zenoss recently came out with a new zenpack so that
they can monitor your OpenStack systems, including things like how many
flavors you have available and how many of each type are in use, how many
images are available and how many are in use, what the various server states
are, etc.... This is in addition to what they can monitor at the individual
server level, if you're also including that as well.

Has anyone played with this new zenpack and can give me some feedback
about how it has impacted your site?

Thanks!

--
Brad Knowles bknowles@ihiji.com
SAGE Level IV, Chef Level 0.0.1