preprod and development are now using the same slaves.. how do we make this work?

RESOLVED FIXED

Status

RESOLVED FIXED
8 years ago
5 years ago

People

(Reporter: dustin, Assigned: rail)

Tracking

Firefox Tracking Flags

(Not tracked)

Details

We are now using the same set of slaves for preproduction and development - developers lock preproduction slaves to development masters while doing their development, then release them back into the preproduction pool.

This has two problems:

1. Currently, preproduction slaves have different passwords from development masters.  This can probably be fixed by just standardizing on the (easy to remember!) preproduction password.

2. The two pools use different SSH keys.  I'm not sure how best to work this out.  We need to:
 a. be able to move slaves between preprod and dev without touching them
 b. maintain whatever security is required in preprod
 c. not depend on configuration differences between preprod and dev

My thinking is that the preproduction SSH destinations (aus, stage, etc.) should be configured to accept the staging SSH keys, and the preproduction keys discarded.  We could also do the opposite, and configure the staging SSH destinations to accept the preprod keys, then get rid of the staging keys.
Rail's fixing this in bug 650007.  I assume folks will figure out the password thing soon enough.
Status: NEW → RESOLVED
Last Resolved: 8 years ago
Depends on: 650224
Resolution: --- → FIXED
Product: mozilla.org → Release Engineering
You need to log in before you can comment on or make changes to this bug.