Testrail status follow-up (ops)

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view

Testrail status follow-up (ops)

Miles Crabill
Hello everyone,

This is a follow-up to Adrian's (great + accurate!) email about the
Testrail incident that occurred today (10/28).

There was some confusion between svcops, Softvision, and the QA groups
regarding the Testrail cutover. The cutover from the old Testrail instance
(managed by sphilp) to the new Testrail instance (managed by svcops)
occurred yesterday (10/27, 13:15 PDT).

I should have done better due diligence and confirmed with Softvision about
the status of the new Testrail instance prior to performing the DNS cutover.

Unbeknownst to me, Softvision had some work missing from the new instance.
A safe time-frame estimate of the missing data is 10/26, 13:15 PDT to
01/27, 13:15 PDT. This morning (10/28, 07:00 PDT) I was alerted about this
missing data, and additionally to an issue where attachments and reports
were inaccessible.

As of 10/28, 08:00 PDT attachments and reports were accessible, and I had
communicated with Adrian from Softvision about the status of the cutover.
We determined that work had been re-done on the new instance, and so a
simple database migration (old -> new) was not advisable. Instead of
committing to a database merge, affected Testrail users are encouraged to
copy missing work from the old instance to the new one.

Both instances of Testrail are currently accessible (from behind the
Mozilla VPN):

- new: https://testrail.stage.mozaws.net
- old: https://testrail-old.stage.mozaws.net

Assuming all else is well, old Testrail instance will be decommissioned in
two weeks (11/14).

Please direct any concerns about the new Testrail instance to me: IRC:
miles, email: [hidden email].

Miles Crabill
dev-quality mailing list
[hidden email]