nsync fails to update desired state in BBS silently

[Amit-PivotalLabs]

Thanks for submitting an issue to capi-release. We are always trying to improve! To help us, please fill out the following template.

Issue

nsync fails to update desired state in BBS silently.

Context

If capi.nsync.cc.base_url is provided without a protocol, e.g. api.system-domain.com instead of https://api.system-domain.com, then rather than failing fast at deployment time, the bulker will continually fail to update freshness because of an invalid scheme "".

Eventually, stale LRPs are left around in Diego and aren't GC'd, leading to insufficient resources.

Steps to Reproduce

Deploy CF + Diego with capi.nsync.cc.base_url set to api.YOUR_SYSTEM_DOMAIN without any http:// or https:// scheme in the URL.

Expected result

When misconfiguring the URL like this, I'd expect to nsync to fail on the deployment, since the configuration it has been given is bad and will never work.

Current result

Currently nsync deploys correctly, and only after a long time do I eventually see lots of app pushes fail due to insufficient resources.

Possible Fix

Validate the input in the startup or pre-start scripts, and fail fast if the input is invalid.

[cf-gitbot]

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

https://www.pivotaltracker.com/story/show/126484591

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

[ameowlia]

Hi Amit,

done! 8296a1a

If this solves your problem, please close this issue.

Thanks, Amelia

[Amit-PivotalLabs]

Haven't tested it out, assuming it works.