691952 - jenkins.mozilla.org, migration

Status: RESOLVED FIXED

(Infrastructure & Operations Graveyard :: WebOps: Labs, task, P1)

Description

[[e-mail] [:onecyrenus]]

Looking to migrate a jenkins job i have running on a qa server  http://10.8.73.228:8080/job/webapps/lastBuild/

Basic instructions on what needs to be installed on the machine is below. 

https://github.com/mozilla/openwebapps/blob/develop/site/tests/scripts/jenkins_install.txt

Basically nodejs, npm a few other requirements, please use the version i have written in the file, and not what is in apt-get.

Comment 1

[Philippe M. Chiasson (:gozer)]

(In reply to [e-mail] from comment #0)
> Looking to migrate a jenkins job i have running on a qa server 
> http://10.8.73.228:8080/job/webapps/lastBuild/

So you are looking for a dedicated jenkins server for openwebapps ? Or are you looking for
a dedicated slave that can be hooked up to jenkins.mozilla.org ?

> Basic instructions on what needs to be installed on the machine is below. 
> 
> https://github.com/mozilla/openwebapps/blob/develop/site/tests/scripts/
> jenkins_install.txt
> 
> Basically nodejs, npm a few other requirements, please use the version i
> have written in the file, and not what is in apt-get.

Does this need to be publicly accessible at all, or is behind the VPN good enough ?

Comment 2

[[e-mail] [:onecyrenus]]

Philippe, 
    I am looking to run my tests on jenkins.mozilla.org, if that means it runs on a dedicated slave, or locally that is up to you. It does need to be publicly accessible. 

Thanks

Comment 3

[[e-mail] [:onecyrenus]]

I should modify that last statement to say the results should be publicly accessible.

Comment 4

[Philippe M. Chiasson (:gozer)]

openwebapps-test1.vm1.labs.sjc1.mozilla.com is up and running, and I installed the requirements you highlighted.

You can login to that VM via the VPN. Your LDAP credentials should let you in via ssh, and you have sudo access to that box.

Not sure what/who is involved in getting this hooked into jenkins, however.

But for now, I'd make sure your test job can successfully run on that VM, that there are no missing dependencies, etc.

Comment 5

[[e-mail] [:onecyrenus]]

So i don't think this is good enough.  The reason i got the other staging server
openwebapps-jenkins.qa.phx1.mozilla.com 


i can take a permanent machine, openwebapps.vm1.labs.. but don't want to go through the hassle of setting up a second instance just to tear it down in the future.

Comment 6

[[e-mail] [:onecyrenus]]

slightly disjointed statement, the reason i got the previous staging server was to iron out the kinks.  

Requirements are changing, so setting up a slave which connects in is the right way to go.  But I don't want to do that just to have it get torn down, and a new configuration needed.

Comment 7

[Philippe M. Chiasson (:gozer)]

(In reply to [e-mail] from comment #6)
> slightly disjointed statement, the reason i got the previous staging server
> was to iron out the kinks.  
> 
> Requirements are changing, so setting up a slave which connects in is the
> right way to go.  But I don't want to do that just to have it get torn down,
> and a new configuration needed.

Not sure what you mean here. Looks like openwebapps-jenkins.qa.phx1.mozilla.com is down at the moment, and I don't know if there is a bug open for tracking that.

Not sure if you need a VM in the labs cluster then, but if you do want it, it's not temporary, and wouldn't be torn down.

Just let me know what you want to do.

Comment 8

[[e-mail] [:onecyrenus]]

ok i'm using it, what is the backup procedure for this machine btw. ? 

Can i have regular snapshots made ?

Comment 9

[[e-mail] [:onecyrenus]]

btw it looks like the virtual machine is not able to connect to the outside world ? 

is this to be expected.. you can login via ldap.. try ping www.yahoo.com or something.. not dns related, possibly routes ? 

netstat -arn reveals routing is fairly ugly .. any thoughts ?

Comment 10

[[e-mail] [:onecyrenus]]

ignore last comment about netstat -arn.. i can reach gateway.. just installed traceroute

Comment 11

[[e-mail] [:onecyrenus]]

traceroute was a no go btw..  Networking connectivity to the internet is down.. unsure why.. no one has told me anything ?

Comment 12

[matthew zeier [:mrz]]

Zandr's time shifted.  Pushing queues around in case someone in server-ops can handle this before he gets online.

Comment 13

[Philippe M. Chiasson (:gozer)]

Should work now, firewall needed some kicking.