Page MenuHomePhabricator

rename cloudvirt1019 and cloudvirt1020 to cloudvirtdb1001 and cloudvirtdb1002
Closed, DeclinedPublic


If @Bstorm can stand this, let's rename these boxes once we have them actually working. Otherwise I'm worried that having some continuously-numbered boxes which are used for totally different purposes will cause endless confusion.

We're doing this same thing with the new data lake boxes, calling them cloudvirtanalyticsXXXX.

Event Timeline

aborrero changed the task status from Open to Stalled.Dec 19 2018, 12:34 PM
aborrero triaged this task as Medium priority.
aborrero added a subscriber: faidon.

Today I had a chat with @faidon and the following concerns were raised about the rename:

It was mentioned that we are already doing a lot of renames (to move from the general prefix labs to cloud), and that already causes additional maintenance burden. In general, renaming is not something we usually do here, and this is an exception already.
I explained to @faidon that this is just our team trying to do better in server naming, that we made some mistakes in the past and that using generic names (like cloudvirt) isn't accurate and creates confusion for us.
Also, I mentioned that these servers have been never into actual service yet due to severe HW issues T196507: Degraded RAID on cloudvirt1019 and T194855: Degraded RAID on cloudvirt1020.

@faidon explained that the key moment to change the naming scheme for servers is when doing HW refresh/procurement. Renaming causes some caos, because there's capex planning spreadsheets, budget spreadsheets, finance's web tools, etc etc.

So the advice is that we don't do this renaming this time.