Change Tracking?

[Ernest Mueller]

Hey all!  I am getting set to try to jumpstart some change tracking at my current gig.  They have JIRA for requests and planning work, but nothing coordinated for actual "I just made a change" tracking besides svn/git commit logs.  I'd like something to cover automated deploys/etc ("Code version 2.3 jut got deployed to staging" or even crons) but also manual changes ("I just restarted that server cause it was locked up with an OOM error").  Without installing a CMDB or some huge software package requiring modification of all kinds of tools and processes.

I think I might have heard of some cool SaaS doohickies that do this nowadays?  My default path will be an email address that scripts/people can use with a semi-formatted subject line and then pull them into Splunk and otherwise try to stick them onto event timelines in Datadog etc. But something that was more "PagerDuty-esque" that one could interact with via API plus mail plus integrate with other stuff more automatically would be nice.  Ideas?

Thanks,
Ernest

[Spike Morelli]

Hi Ernest,

SaaS wise, you could plug your events into Librato Metrics as Annotations: http://dev.librato.com/v1/annotations . Especially if you sent them metrics too then it gets pretty powerful. They have a nice API

Alternatively Graphite supports events too stored in a django db. I can't find the docs to the API now and don't recall what kind of granularity you have in terms of interacting with them, but shouldn't be too shabby.

Last and somewhat less orthodox approach is to run your run twitter like server using Identica and treat all those things as tweets (Identica lets you adjust the 140 chars). This is actually pretty interesting, rich API and lots of fun stuff you can make with (iirc @ripienaar experimented with something similar a while back with some nice results but don't recall his conclusions). Personally it worked well for me and some folks, we'd ship a little cli on all servers so that we could just do "lart 'rebooting this for OOM'" and see it in a stream of the team or intermixed with "tweets" from boxes about their cronjobs and whatnot.

hope that helps,

Spike

--
You received this message because you are subscribed to the Google Groups "devops-toolchain" group.
To unsubscribe from this group and stop receiving emails from it, send an email to devops-toolcha...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

Spike

--

http://www.spikelab.org/

http://twitter.com/spikelab

http://www.linkedin.com/in/spikemorelli

[Aaron Bull Schaefer]

On Tue, Mar 19, 2013 at 11:01 AM, Ernest Mueller
<ernest....@gmail.com> wrote:
> I'd like something to cover automated deploys/etc ("Code version 2.3 jut got
> deployed to staging" or even crons) but also manual changes ("I just
> restarted that server cause it was locked up with an OOM error"). Without
> installing a CMDB or some huge software package requiring modification of
> all kinds of tools and processes.

I've been meaning to install and test out Anthracite [1] for this use
case. It's described as "an event / change management app to enrich
monitoring dashboards and to allow visual and numerical analysis of
events that have a business impact". Looks to be fairly flexible in
how you add and query events.

[1] https://github.com/Dieterbe/anthracite

--
Aaron Bull Schaefer

[Vladimir Vuksan]

Problem here is what and how you want to correlate it. Inserting events
into a data store is easy enough however the tricky part is what to do
with afterwards and this depends on what other tools you are using. People
have mentioned annotations and events support for Librato and Graphite. We
use Ganglia for most of our metrics and trending so we tend to add events
into it ie. we have a command line tool that you could say things like

logit testing network settings

and an event will be added via Ganglia API to that host. I have also
modified init scripts in the past to add things like "Jira restarted by
init script" to denote intentional restarts vs. app crashes.

Vladimir

[Peco Karayanev]

I think there is more to change tracking than events/stats which capture a change after it is implemented. I think Ernest may be looking for overall change management, which includes the request/planning/approval states.

 

thanks

--
You received this message because you are subscribed to the Google Groups "devops-toolchain" group.

To unsubscribe from this group and stop receiving emails from it, send an email to devops-toolchain+unsubscribe@googlegroups.com.

[Ernest Mueller]

Thanks all! I like the librato and anthracite ideas, although I already have loads of visualization and other tools (splunk zabbix datadog nagios custom webapps) so want to keep it lean and feed change events to those instead of adding another UI.  Anthracite with mods to take submission via emails for quick integration with random crons might work though.

--
Aaron Bull Schaefer

--
You received this message because you are subscribed to the Google Groups "devops-toolchain" group.

To unsubscribe from this group and stop receiving emails from it, send an email to devops-toolcha...@googlegroups.com.

[Ernest Mueller]

Nah, I want to start simple. We more-or-less use JIRA for change approval. (Though I do have some clever startup-y ideas for more flexible request/approval stuff I've schemed with Peco about, this isn't part of that.)   We have zip here.  At NI we had the mail-in database, which was clearly better than nothing and has a extremely low barrier to entry from both humans and automated processes, so I'm trying to figure out if there's anything spiffier with an activation energy somewhere around that.

To unsubscribe from this group and stop receiving emails from it, send an email to devops-toolcha...@googlegroups.com.

[Ajey Gore]

I personally like droptask

--
Thanks

Ajey

[jyang]

My startup (www.metaforsoftware.com) has created a light-weight tool to track changes (both expected and unexpected ones).  It can be scheduled to check for server changes hourly or daily and get email notification when changes are detected.  Can also be used to detect drift across cloned servers in a cluster.  Also, we have APIs but haven't made it public yet (need some clean up).  We're in free beta right now.  Would love to get your thoughts on it and can get you access to APIs depending on your need. 

On Monday, March 25, 2013 1:32:39 AM UTC-7, Ajey Gore wrote:

I personally like droptask

On Wed, Mar 20, 2013 at 2:04 AM, Ernest Mueller <ernest....@gmail.com> wrote:

Nah, I want to start simple. We more-or-less use JIRA for change approval. (Though I do have some clever startup-y ideas for more flexible request/approval stuff I've schemed with Peco about, this isn't part of that.)   We have zip here.  At NI we had the mail-in database, which was clearly better than nothing and has a extremely low barrier to entry from both humans and automated processes, so I'm trying to figure out if there's anything spiffier with an activation energy somewhere around that.

On Tue, Mar 19, 2013 at 3:28 PM, Peco Karayanev <pecom...@gmail.com> wrote:

I think there is more to change tracking than events/stats which capture a change after it is implemented. I think Ernest may be looking for overall change management, which includes the request/planning/approval states.

 

thanks

On Tue, Mar 19, 2013 at 4:03 PM, Vladimir Vuksan <vli...@veus.hr> wrote:

Problem here is what and how you want to correlate it. Inserting events into a data store is easy enough however the tricky part is what to do with afterwards and this depends on what other tools you are using. People have mentioned annotations and events support for Librato and Graphite. We use Ganglia for most of our metrics and trending so we tend to add events into it ie. we have a command line tool that you could say things like

logit testing network settings

and an event will be added via Ganglia API to that host. I have also modified init scripts in the past to add things like "Jira restarted by init script" to denote intentional restarts vs. app crashes.

Vladimir

On Tue, 19 Mar 2013, Ernest Mueller wrote:

Hey all!  I am getting set to try to jumpstart some change tracking at my current gig.  They have JIRA for requests and planning work,
but nothing coordinated for actual "I just made a change" tracking besides svn/git commit logs.  I'd like something to cover automated
deploys/etc ("Code version 2.3 jut got deployed to staging" or even crons) but also manual changes ("I just restarted that server cause
it was locked up with an OOM error").  Without installing a CMDB or some huge software package requiring modification of all kinds of
tools and processes.

I think I might have heard of some cool SaaS doohickies that do this nowadays?  My default path will be an email address that
scripts/people can use with a semi-formatted subject line and then pull them into Splunk and otherwise try to stick them onto event
timelines in Datadog etc. But something that was more "PagerDuty-esque" that one could interact with via API plus mail plus integrate
with other stuff more automatically would be nice.  Ideas?

--
You received this message because you are subscribed to the Google Groups "devops-toolchain" group.

To unsubscribe from this group and stop receiving emails from it, send an email to devops-toolcha...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

--
You received this message because you are subscribed to the Google Groups "devops-toolchain" group.
To unsubscribe from this group and stop receiving emails from it, send an email to devops-toolcha...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

--
You received this message because you are subscribed to the Google Groups "devops-toolchain" group.
To unsubscribe from this group and stop receiving emails from it, send an email to devops-toolcha...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

--
Thanks

Ajey