Running out of ephemeral disk on GCP about cf-deployment

[dsabeti]

@ljfranklin Our GCP does a fresh deploy every time (it deletes the existing deployment first), so it's possible that we'd never see this problem. I'll prioritize this story to bump up to 10GB.

[dsabeti]

Here's the story to follow: https://www.pivotaltracker.com/story/show/136972395

[dsabeti]

All of the stories associated with this have been accepted, so I'm going to close this issue.

[cf-gitbot]

We have created an issue in Pivotal Tracker to manage this:

https://www.pivotaltracker.com/story/show/140105709

The labels on this github issue will be updated when the story is started.

[dsabeti]

@ljfranklin How often do you see this problem? I'm not sure that we've seen this problem in our GCP pipelines. Also, is this in a fresh deploy or rolling/upgrade deploy?

@anEXPer, have you seen this problem in our builds before?

[ljfranklin]

@dsabeti only saw it once so far and it was during a rolling deploy. I'd imagine others would see it on jobs with very chatty logs or in our case a rolling deploy with large packages to untar. A 1GB ephemeral partition is bound to run out of space at some point if large packages need upgraded.

[dsabeti]

Hey @ljfranklin. master now contains 10GB_ephemeral_disk for everything that has ephemeral disk attached.

[ljfranklin]

@dsabeti we're still seeing this issue in CI as the operations/gcp.yml ops file still overrides everything to 5GB. Do we even need that ops file or can it be removed?

[dsabeti]

Yup, it looks we should get rid of those extra vm_extensions in that ops file.