Closed Bug 727353 Opened 13 years ago Closed 13 years ago

New Selenium Hub with 2 nodes in phx1

Categories

(Infrastructure & Operations :: Virtualization, task)

task
Not set
normal

Tracking

(Not tracked)

RESOLVED FIXED

People

(Reporter: fox2mike, Assigned: dparsons)

References

Details

+++ This bug was initially created as a clone of Bug #693544 +++ Dave Hunt (:davehunt) 2012-02-14 03:17:45 PST We should be able to set up a Selenium Grid instance for Jenkins to run tests against with the following: 1x Linux instance - to run Selenium Grid hub on. 2x Windows 7 Professional VMs - to run the Selenium server nodes and the actual tests. This will allow us to run tests in parallel, and we should be able to scale up in future by adding more VMs. Ultimately we may scale the Phoenix grid up to replace the grid in MV. Marlena: Could you provide Dan with a vmdk from one of our virtual machines? We would need to be able to access the Linux box via SSH, and be able to install Java, Ant, and Git on there. We would need to be able to access the Windows VMs via remote desktop or similar. A suggestion for the hostnames: * selenium-hub (Linux box) * selenium-node1 * selenium-node2
Can we call these machines : selenium-hub1.dmz.phx1.mozilla.com selenium1.dmz.phx1.mozilla.com selenium2.dmz.phx1.mozilla.com Thanks! Stephen/Dave, Can you please give the hostnames of the machines running the vmdks, so Dan can pull what he needs and get this going? Thanks!
I've selected as candidate and shut down the Windows VM on the Mac machine qa-selenium6.mv.mozilla.com
Zac, Awesome. Can you email Dan with the access credentials? Also, is this a selenium hub or node?
I have emailed Dan already :) It is a node.
Zac, can you please give me the full path to the directory that contains the VM you want me to move from qa-selenium6.mv.mozilla.com? I logged in and found two different places where VMs were stored and three different Windows VMs and I was unable to tell exactly which VM you want transferred.
Assignee: server-ops → dparsons
Dan, sorry about that. The one we want is at: /Users/retornam/Virtual Machines/Windows7SP1.20111122.vmwarevm
The VM has been copied to fs2 in mtv1 so it could be booted up again on qa-selenium6. It's now being copied to phx1, which will take a while, and then it needs to be converted. I'll let you know when it's done.
Are these up and running? Wondering if we can access these yet.
Dan is at scl3 setting up the new ESX infrastructure, so he may not have gotten this fixed up yet. A lot of things are blocked on that new infrastructure, so if it's OK by you to hold out a few more days, that would be great.
(In reply to Dustin J. Mitchell [:dustin] from comment #9) > Dan is at scl3 setting up the new ESX infrastructure, so he may not have > gotten this fixed up yet. A lot of things are blocked on that new > infrastructure, so if it's OK by you to hold out a few more days, that would > be great. Sure thing; can we get insight into that ETA? I know there is work being done this week; is there another tracking bug we can follow? Again, this is crucial to getting Webdev+WebQA working together at another level, naturally.
Can the email with access info be sent to me again, and also can someone please summarize the request into one comment that specifically answers: (1) FQDN for each VM (2) RAM, CPU and disk space needed for each VM (3) What I need to copy from where, or if some of these VMs should be new, etc
VM1: 1: selenium-hub1.dmz.phx1.mozilla.com 3: New VM running Linux VM2: 1: selenium1.dmz.phx1.mozilla.com 3: Copied from fs2 as mentioned by lerxst in comment 7 VM3: 1: selenium2.dmz.phx1.mozilla.com 3: Copied from fs2 as mentioned by lerxst in comment 7 Ideally we would shutdown our VM again for it to be copied, but let me know if you need to do this again. I've intentionally not specified RAM, CPU, and disk space. The existing Windows 7 VMs have 1 core, 1GB RAM, and 40GB HDD. This would be a minimum for all VMs, but I am not aware of what is available. I hope this helps.
Copying this from mtv1 to phx1 now. It's taking an unusually long time.
I've spent the last few hours trying to convert this VM and it's not going to work because of the many, many, many snapshots you have running on the VM. This is what you need to do: (1) Shut down the VM (2) Remove absolutely every single snapshot you have running (3) Copy the VM to somewhere on fs2 (4) Once the copy is complete, you can start the VM back up on your end. Update this bug when the copy is complete, along with where you put the VM files, and I'll start it copying to phx1.
I'm surprised that there's more than a couple of snapshots on that VM to be honest. Would it be easier to go with a fresh installation of Windows 7 Pro? We have only a few changes that we'd need to make. We could even set one of them up and when it's done clone it for the second VM?
Yes, that would be a lot easier at this point. I'll start working on that.
Two questions: (1) For selenium-hub1.dmz.phx1.mozilla.com, is RHEL 6 an acceptable OS? (2) Also for selenium-hub1.dmz.phx1.mozilla.com, I need a list of LDAP usernames that need SSH access, along with whether or not they need sudo access. For "LDAP username" I'm looking for, as an example, "dparsons" for myself, as the username I use to log into say, zimbra, is "dparsons@mozilla.com". And remember, for each username that needs to be able to ssh in, please tell me if they also need sudo access. Unfortunately I can't work on this bug until the above questions are answered.
(In reply to Dan Parsons [:lerxst] from comment #17) > Two questions: > > (1) For selenium-hub1.dmz.phx1.mozilla.com, is RHEL 6 an acceptable OS? Yes, that should be fine. > (2) Also for selenium-hub1.dmz.phx1.mozilla.com, I need a list of LDAP > usernames that need SSH access, along with whether or not they need sudo > access. For "LDAP username" I'm looking for, as an example, "dparsons" for > myself, as the username I use to log into say, zimbra, is > "dparsons@mozilla.com". And remember, for each username that needs to be > able to ssh in, please tell me if they also need sudo access. sdonner, zcampbell (sudo), dhunt (sudo), mbrandt, retornam (sudo). That should be enough for now; we can always change this list later, right?
And, sorry, one more addition: mcompton; thx.
zcampbell will not have access until he emails me his ssh public key.
This is all done, except for zcampbell's access. I'll set that up when he emails me his ssh pub key. You may need to ask netops for flows to allow you to remote desktop in or ssh in. selenium-hub1.dmz.phx1.mozilla.com selenium1.dmz.phx1.mozilla.com selenium2.dmz.phx1.mozilla.com
Status: NEW → RESOLVED
Closed: 13 years ago
Resolution: --- → FIXED
(In reply to Dan Parsons [:lerxst] from comment #21) > This is all done, except for zcampbell's access. I'll set that up when he > emails me his ssh pub key. > > You may need to ask netops for flows to allow you to remote desktop in or > ssh in. > > selenium-hub1.dmz.phx1.mozilla.com > selenium1.dmz.phx1.mozilla.com > selenium2.dmz.phx1.mozilla.com Raymond filed 741956.
I have started a Selenium Grid hub on selenium-hub1.dmz.phx1.mozilla.com and a node on selenium1.dmz.phx1.mozilla.com. Feel free to send tests to it, the environments available are Firefox 12 and Internet Explorer 8 on Windows 7. A few tweaks were needed for moz-grid-config, which we will need to either maintain local patches for, add some logic for, or maintain a separate branch for. Also, if these VMs suffer any downtime or the Selenium servers fail for any reason we will need to login and restart them. To do this on the hub: 1. SSH to selenium-hub1.dmz.phx1.mozilla.com 2. cd ~/moz-grid-config 3. ant launch-hub To do this on the node: 1. Remote Desktop to selenium1.dmz.phx1.mozilla.com 2. Start a command prompt 3. cd /Users/mozwebqa/moz-grid-config 4. ant launch-node -Dhub.host=selenium-hub1.dmz.phx1.mozilla.com -Dnode.host=selenium1.dmz.phx1.mozilla.com We should also look into getting this to start automatically on boot for these VMs.
Product: mozilla.org → Infrastructure & Operations
See Also: → 1112637
You need to log in before you can comment on or make changes to this bug.