Viewing profile — imbradn

It was carried forward from days before checksums were a thing for the main data files. Enabling them on existing clusters was not added until PG 14 (in core - there was a utility …

We used it to shard a database at Instacart after your time instead of the “replicate everything and drop the data on the shard approach”. That combined with dropping all the index…

Logical replication row filtering is a much better alternative to the ship and drop approach.

Often lost in these discussions is how much more difficult upgrading is at scale. The article talks about the challenges with upgrading a 4TB database. In my world - that’s a small…

No, vacuum issues are not solved. This will reduce the amount of scanning needed in many cases when vacuuming indexes. It will mean more efficient vacuum and quicker vacuums which …

It's based on Patroni - which they wrote and maintain and is the most popular HA Failover software for Postgres.

The acquisition of Oriole is very interesting. Getting that to a GA product would be a game changer for a lot of workloads.

Except for the un-implemented features which they might need. It also uses serializable isolation and in their implementation reads are blocked by writes unlike in Postgres. Those …

IOPS isn’t a linear thing here. The vacuums needed to prevent transaction wraparound (vaccum Freeze)can’t be throttled and are much more expensive than regular vacuums. By splittin…

That article covers the basis of how we do upgrades at Instacart, but is quite old. This is a more modern look at how we accomplish the process. We have used this process to upgrad…

But in those days, in addition to the license prices for closed source DBs, you also had to pay all the hardware costs to run them on, and hire an army of DBAs to support and maint…

There is a History tab with examples, such as: https://explain.depesz.com/s/vzlS This is an indispensable tool for Postgres query plan analysis. Even for those who are experienced …