Re: [PERFORM] How we made Postgres upserts 2-3* quicker than MongoDB

2016-01-08 Thread Mark Zealey
On 08/01/16 19:07, Nicolas Paris wrote: Hello Mark, As far as I know, MongoDB is able to get better writing performances thanks to scaling (easy to manage sharding). Postgresql cannot (is not designed for - complicated). Why comparing postgresql & mongoDB performances on a standalone instance si

[PERFORM] How we made Postgres upserts 2-3* quicker than MongoDB

2016-01-08 Thread Mark Zealey
Hi all, I just wrote an article about the postgres performance optimizations I've been working on recently especially compared to our old MongoDB platform https://mark.zealey.org/2016/01/08/how-we-tweaked-postgres-upsert-performance-to-be-2-3-faster-than-mongodb -- Sent via pgsql-performance

Re: [PERFORM] Proposal for unlogged tables

2016-01-04 Thread Mark Zealey
On 04/01/16 18:12, Andres Freund wrote: Pages containing data of unlogged tables aren't ever flushed to disk unless a) a shutdown checkpoint is performed b) a buffer containing data from an unlogged table is used for something else c) the database being copied is the the source of a CREATE

Re: [PERFORM] Proposal for unlogged tables

2016-01-04 Thread Mark Zealey
On 04/01/16 16:27, Yves Dorfsman wrote: I haven't tried, but won't converting an unlogged table into a logged table write all the inserts at once instead of once per insert? Or are you wanting to do more bulk insert into that table later? Are you trying to avoid running a CHECKPOINT? Are you

[PERFORM] Proposal for unlogged tables

2016-01-04 Thread Mark Zealey
I've recently been doing some performance testing with unlogged tables vs logged tables on 9.5-rc1. Basically we're trying to do big loads of data into the database periodically. If on the very rare occasion the server crashes half way through the import it's no big deal so I've been looking sp