-
Notifications
You must be signed in to change notification settings - Fork 167
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
ansible: add 1015 orka test macs #2253
Conversation
ansible/inventory.yml
Outdated
@@ -153,6 +153,9 @@ hosts: | |||
macos10.11-x64-2: {ip: 207.254.58.162, port: 10004, user: administrator} | |||
macos10.12-x64-1: {ip: 207.254.58.162, port: 10001, user: administrator} | |||
macos10.12-x64-2: {ip: 207.254.58.162, port: 10002, user: administrator} | |||
macos10.15-x64-1: {ip: 199.7.167.99, port: 8823, user: administrator} | |||
macos10.15-x64-2: {ip: 199.7.167.100, port: 8823, user: administrator} | |||
macos10.15-x64-3: {ip: 199.7.167.99, port: 8824, user: administrator} |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM
Maybe we should brand these in Jenkins as "Orka", so |
Ive managed to sort out the 10.14 issues and get a working VM which I will now create an image of. Ive redployed all machines to remove an 1015 machine (thought 3 here was a bit overkill) and to turn off VNC access on the VMs. |
Ive added a release machine, reansibled to get the correct hostname and installed full xcode on it. The remaining release setup is:
|
@nodejs/build-infra whats the process about adding IP to the jenkins firewall? is just the IP of the machine needed or the ports aswell? If its just the IPs does it make sense to add all three orka IPs to both jenkins hosts to make it easier to add new machines at our whim? |
@AshCripps just the IP addresses, and yes, we should just add all 3 I reckon. That'd be 199.7.167.99, 199.7.167.100 and 199.7.167.101? |
@rvagg yes please those 3. Ill try to create some jenkins instances for test today. |
done, added entries:
|
test ci build - https://ci.nodejs.org/job/node-test-commit-osx-AshCripps/11/ EDIT: xcode got corrupted on 10.15 I think so redownloading before I rerun the build. |
@rvagg @nodejs/build I think we need to rethink our migration of our vsphere systems. ATM we only need osx10.10 for the building of release sources (#2282) so I tried to pull down an image of the test machine today. Not only is there a minefield you have to go through due to the outdatedness of the vsphere when i finally started the download it was going to take over a year according to my rough maths! So IMO that option is out and leave us with a couple of options:
LMK your thoughts |
We switched release source building from macOS to Linux for 12+ in #1777 so if we were to migrate away from macOS 10.10 it would make more sense to use the same Linux nodes as 12+ rather than attempt to go to another version of macOS. |
Sounds like getting #1777 back onto 10.x would be the best way, and also simplify (hopefully) the range of build variations we have. |
Is it worth considering hanging on to 10.10 in the old vsphere stuff until April 2021 to see out v10.x? The other options are probably not high risk. We originally switched to linux for tarballs because bsd tar was producing tarballs that weren't as compatible as gnu tar, so we should increase compatibility by switching to linux for the sources files. I haven't heard of any issues with the switch between 10 to 12, only the opposite - complaints went away. You never know what weird edge cases people out there are relying on that we might trigger, but we could make a switch like we did for notarization and just include a big notice in the next 10.x that they've changed but we don't expect problems. |
IMO I dont think its worth, seems a bit silly to hold on to such aging hardware for only 1 machine and macstadium would like to get rid of it sooner rather than later. |
Hoping to add the 1015 and 1014 machines to the CI jobs today. Still need someone to make the release CI agent and do the final setup steps for the second 1015 release machine EDIT: all build passed except 12.x on 1015 which failed to a known flaky test which is in process of being removed - nodejs/node#32870. Im going to add these machines to the CI jobs now |
@@ -47,7 +47,7 @@ | |||
# providers - validated for consistency | |||
'provider': ('azure', 'digitalocean', 'ibm', 'joyent', 'linuxonecc', | |||
'macstadium', 'marist', 'mininodes', 'msft', 'osuosl', | |||
'rackspace', 'requireio', 'scaleway', 'softlayer', 'voxer', | |||
'orka', 'rackspace', 'requireio', 'scaleway', 'softlayer', 'voxer', |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
this is breaking master since the orka hosts were merged into secrets
Exception: Failed to parse host: Invalid provider: orka
I can fix manually to run Ansible but can we get this merged and have follow-up work later?
Finally got the firewall set up so heres the draft PR whilst I finish setup.
Couple things need doing.