Re:Zero downtime Postgres migration, done right
by scrollaway on Friday June 11, 2021 @12:38PM
(#27475473)
I remember reading something like that about Adyen. That might be why you're unable to find it.
Re:Zero downtime Postgres migration, done right
by rattray on Friday June 11, 2021 @04:13PM
(#27477881)
I wonder if that was it (Adyen and Braintree are competitors).
Here's the post: https://www.adyen.com/blog/updating-a-50-terabyte-postgresql...
Discussed here: https://news.ycombinator.com/item?id=26535357
They architected their application to be able to tolerate 15-30min of postgres downtime.
Re:Zero downtime Postgres migration, done right
by nezirus on Friday June 11, 2021 @01:02PM
(#27475811)
pgbouncer has PAUSE comnand, which can be used for seamless restarts/failovers or similar
Re:Zero downtime Postgres migration, done right
by simonw on Friday June 11, 2021 @03:21PM
(#27477400)
I think that's covered in this talk (I've not watched the video though): https://www.braintreepayments.com/blog/ruby-conf-australia-h...
Re:Zero downtime Postgres migration, done right
by pgr0ss on Friday June 11, 2021 @03:38PM
(#27477545)
It was also covered in: https://www.braintreepayments.com/blog/switching-datacenters...
This app apparently had robust enough retry mechanisms that they were able just eat the errors and not have customer issues—Color me impressed! I'm not sure how many teams can make that claim; that's a hard thing to nail down.