Running a own Kadira instance Update: now with a guide!

Hello,
UPDATE: https://medium.com/@foodfitnesscode/setting-up-kadira-on-a-ubuntu-16-04-server-part-1-a821cdddd1e6

so the latest info is that kadira will Shutdown on 15. Mai.
We are currently using kadira.io and so I was trying to run kadira on our own server.

I cloned https://github.com/kadira-open/kadira-server
I was trying to build the docker container with docker-compose build. This Failed.
So I was trying to run everything directly after some hours I stopped frustrated.
Then I found https://github.com/kadira-io where are no dockerfiles anymore and there is no documentation how to run kadira.

Will there be docker container published until Mai 15?
Or will there be a guide how-to setup Kadira ?

@sashko @benjamn

12 Likes

I would really appreciate also an answer on this topic

1 Like

Me too. I would like to learn more how to do that.

Unfortunately @sashko mentioned in some posts and a youtube video that the MDG staff won’t reply to any requests regarding the self-hosting of Kadira APM.
So we gotta figure it out ourselves. But I am definitely with you that someone familiar with this topic would help the community a lot by writing just a couple of lines of what to do to self host it.

2 Likes

I still don’t get the MDG move
So they buy the unique APM for meteor, they open source it but without any instructions how to run it and they offer the service only to Meteor Galaxy users?

Why don’t they do something just like Optics?
A service everybody can use even if you are not using Galaxy?
I don’t really get it…

And at least some help or a docker container to easily deploy our own instance of Kadira… but nothing.
Amazing move

@arunoda can you maybe help us. The last request from the community :blush:

5 Likes

I got kadira running localy !

If there is enough interest i can write up a short tutorial.

There are some Databases missing which you have to create by hand etc etc…

10 Likes

Yes, that would be great - step-by-step instructions on what to do would be nice.

I am very interested in this, the move by MDG is really incomprehensible. I just fail to understand why they did not put a couple man-hours into making this even remotely more easy for the community. Do they not understand that some of their users can simply not, under almost any condition, use galaxy…?

@lampe I’d be very thankful if you could help! :slight_smile:

Hey,

I am the project owner of https://github.com/kadira-io, I am knowledgeable of how the APM runs and works, kadira.io is a place for the community to try get involved in stripping out areas of the apm that are no longer relevant, re-branding, documentation and UI refresh setup docker

If you are interested in then drop me a message.

Regards

6 Likes

yes please! :slight_smile:

(My guess would be that the ‘stress’ on the DB would be not that big if people deploy it for their own sites only…?)

MDG did presumably pay Arunoda for it, and then open sourced it immediately, that’s not bad. Just wait a week or two. There’s already a working docker image, now it’s just a matter of cleaning things up.

1 Like

A working docker image? can you point me to that image ?

Ah it’s not public yet I guess, but Robfallows mentioned it here:

A short updated:

I had some time to patch stuff in kadira-server and I want to release the repo this weekend.

Also I’m planning to release blog posts about how to setup kadira on a ubuntu 16.04 server.

12 Likes

Part 1 is up on Medium:

Part 2 coming ASAP

19 Likes

I just released part 2 !

in part 3 i will only give some tups and tricks :wink:

11 Likes

Awesome stuff, thanks for the post!

Wow thanks a lot. Awesome that you wrote this all down!

Awesome, thanks! :smiley:

Hmm, in Part one you say “Clone the repo from”

git clone git@github.com:lampe/kadira-server.git

Then in part two you say “Clone Kadira Server from GitLab”

git clone git@quantilope.githost.io:kadira/server.git

But that one asks me to provide git@quantilope.githost.io’s password :confused:

Also, what is this about Amazon Web Services? :confused:

Sry it was pointing to our internal repo i changed it.
But you don’t need it to clone the second time.

the cpu profiler feature needs s3 bucket where it saves the json.