GitLab Monitor Stage - Speed Runs, 4 Aug 2021

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Rootly.io Speed Run

Description

Rootly.io is a SRE incident management solution inside Slack. This is a speed run of first impressions by GitLab Senior Product Manager, Alana Bellucci.

https://rootly.io/

A

Hi everyone- I am the product manager for the monitor stage here at git lab, and I wanted to talk to you today about rootley. They are a company that is building out an srd incident management solution directly inside slack.

A

Kevin brought this to me my attention and I kind of wanted to take a little bit of a dive and do a quick, speedy run, but they just raised their seed round and they have an interesting product. So, let's dive in and take a look on their home page. They have this really nice gif.

A

That kind of talks about and demonstrates what they're able to do directly within slack to resolve, resolve incidents. um A couple different integrations that I think I've note, jira zoom pagerduty um seems to feel really fairly similar to what some of our homegrown tools are like. So this is woodhouse that our sre team has built out and I'm going to just do a staging incident directly in slack. So if I enter in woodhouse staging incident declare I pop up with something like this. I can and I can create an incident directly in slack um page.

A

You know the respective people that I think are necessary and um this will generate you know an issue in gitlab, a zoom link and then its own slack channel. So that way the incident can be mitigated, but let's go back to rootley.

A

So taking a look at rootly, um they have a lot of different things um that are built out. Some are similar. I I haven't set up the configuration directly for slack, so I'm directly in the app right now, but they have incidents and incidents, look very similar to the incidents that we have started to build out directly in gitlab. They have similar things like that. We're considering doing kind of like incident timelines.

A

Some links over here to you know assume a zoom meeting if you wanted to have some type of war room during the incident- and they also have this interesting thing called post-mortem. So what happened? um What happened during the incident- and why did it happen? um We also have some of these things directly in the incident template again. We see timelines, so this could be another use case for building out those timelines.

A

They do allow you to download download the pdf, which is nice if you need to distribute it amongst team members that maybe don't have access to the incident tasks and action items um tasks I think come from playbooks and um action items, I'm not entirely sure how those differ from action items, but um something to kind of just be aware of genius is a workflow for based on looks to be based on the metadata for an incident.

A

So if something is a certain severity, then do this, and this keeps track of how many times it was run pulses. I think this might be similar to something a feature that we have in git lab, but what has changed?

A

What has kind of changed in the infrastructure environment? So um was there a change in a docker image that recently got deployed? Was there a commit directly on github or git lab? Maybe something different happen within a kubernetes cluster? I don't you know having that visibility.

A

um So that way, you see what maybe recently changed within your environment so that you can quickly track down an incident. Metrics seem to be not necessarily metrics of.

A

What's within your system like prometheus or something that we would have equivalent to the metrics within the tab in gitlab, but more around what is happening directly within those incidents? How did they get resolved? How did they get mitigated and then also some things that are unique to rootley? So, um what's going on with the different post-mortems?

A

Now, let's dive in and just take a quick look at the configuration, so alerts um alerts. I think you have the option to pull in alerts from pagerdutyopsgenie.

A

This is something that I'd like to follow up and take a little bit of a deeper dive in because I hadn't I hadn't heard of this company before.

A

And then going back to configuration, I think um environments so which different environments are these alerts coming from severities? I think this is where you can set what what different severities mean.

A

Services so not totally sure how services are necessarily different than your environments. It looks like maybe it's broken down a little bit like what is your elasticsearch and production doing versus maybe or you know your two different databases or postgres um functionalities?

A

I think this is specific functionalities within whatever app or system their infrastructure team is working on.

A

Playbooks these are, um I think, I mentioned this when we were going over incidents. I think these are um the these are what generate the different tasks for a specific type of incident? Okay, how can you kind of have that different checklist with those different texts, as things come through and make that a repeatable process teams um who are who are the different teams- and you can compare that directly directly in here, postmortem templates?

A

Let's take a look.

A

Okay, so this would be like an issue, an issue type, an incident issue, type template that we could add directly within git lab.

A

Incident pauses, so I think this is a way for teams to map different incident causes to.

A

To whatever the incident is and incident types slightly different um is this: is this cloud? Is this on-prem? Is this customer facing? Is this security related it's kind of breaking that down a little bit more incident rules? I think this is similar to some of the roles that we have for different people that would be on the on-call rotation and then they also have a section for status page.

A

So I think right now we're using status.io, but we have some options to integrate with that, but um yeah still figuring things out here at git lab but um wanted to kind of dive in and see what some other companies were doing. Thanks for watching.