Zabbix / Stackstorm Integration

[brad.tu...@gmail.com]

Hi Guys,

I just heard about Stackstorm.  I'm very interested in it.  I see that stackstorm integration is possible with Zabbix.  I've been using Zabbix for about 8 years, so I'd consider myself a power user, if not expert.

Can someone tell me what Stackstorm does that Zabbix either doesn't do, or doesn't do well?  I know, and am using, the zabbix actions upon events happening, so I'm not convinced that this is a duplication of functionality, albeit in a different form.

Additionally, we are using Ansible, so if someone can touch upon the ability of stackstorm to integrate with ansible and zabbix for a perfect tri-fecta I'd be very greatfull.

We are an ITIL shop, so we're heavily starting to implement Event management.  I have a feeling this solution may be beneficial in achieving a higher level of ITIL conformity.

Thanks,

Brad Turnbough
Zabbix 2.0.12

[Dmitri Zimine]

Hi Brad, thanks for writing, 

I’ll give you a detailed answer later today. The short is StackStorm closes the loop between monitoring and actions, what in ITIL work called “closed loop incident management”. 

Cheers, Dmitri. 

--
You received this message because you are subscribed to the Google Groups "StackStorm" group.
To unsubscribe from this group and stop receiving emails from it, send an email to stackstorm+...@googlegroups.com.
To post to this group, send email to stack...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/stackstorm/8e403733-4b09-4c43-81f5-c7e474aeba99%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Dmitri Zimine]

Brad, 

I’d say, StackStorm is a generalization of your solution, Zabbix + Ansible, to heterogeneous environment :) 

StackStorm is purpose build integration. IFTTT for operations. Get a trigger, apply rule, fire a workflow, pass payload data down from event to action to other action. 

Yes you can fire Ansible playbook when Zabbix detects and triggers a problem.

However, Zabbix, or any monitoring tool, is not the only source of events. 

Git commits, Jenkin Job completes, JIRA ticket updates, Ansible playbook fails… they are all events.

Likewise, there are multiple systems, tools, custom scripts to be called.

Wiring them directly gets very messy very fast.

StackStorm is a control panel for this integration. 

Our design goal is to make it easy to build integrations with sources of events and systems/tools/existing scripts.

The actions are chained into workflows, passing data down the flow (similar to playbook, I recently blogged about it)

Than we wire events with actions via rules - when this happens, and criteria matches, fire an action, or a workflow.

Zabbix integration is writing a Sensor (piece of python code) to listen for trigger notification or pull for problems or raw data… 

We haven’t done it yet; if you got interested to contribute, we can guide you and help. 

Ansible integration is already on the roadmap. 

Once integration packs done, Zabbix->StackStorm->Ansible integration will be just like current Nagios->StackStorm->Salt or NewRelic->StackStorm->Chef. Or mix and match. Or add steps - Ticketing for ITIL complience, Chatops for dynamics, etc. 

Hope I didn’t jump too high on my high horse :) We can get into details. Ask more. 

Cheers, DZ.