Re: badly scaling performance with appending to bytea

2018-03-21 Thread Gary Cowell
;ll see if it's possible to do this on AWS PostgreSQL RDS service but this sort of thing is usually not On 21 March 2018 at 12:59, Pavel Stehule wrote: > > > 2018-03-21 13:56 GMT+01:00 Pavel Stehule : >> >> >> >> 2018-03-21 13:03 GMT+01:00 Gary Cowell : >>

badly scaling performance with appending to bytea

2018-03-21 Thread Gary Cowell
We are trying to implement postgresql code to load a large object into a postgresql bytea in chunks to avoid loading the file into memory in the client. First attempt was to do update build_attachment set chunk = chunk || newdata ; this did not scale and got significantly slower after 4000-5000