On 10/16/24 21:37, Andy Hartman wrote:
> I am very new to Postgres and have always worked in the mssql world. I'm
> looking for suggestions on DB backups. I currently have a DB used to
> store Historical information that has images it's currently around 100gig.
>
> I'm looking to take a monthly backup as I archive a month of data at a
> time. I am looking for it to be compressed and have a machine that has
> multiple cpu's and ample memory.
>
> Suggestions on things I can try ?
> I did a pg_dump using these parms
> --format=t --blobs lobarch
>
> it ran my device out of storage:
>
> pg_dump: error: could not write to output file: No space left on device
>
> I have 150gig free on my backup drive... can obviously add more
>
> looking for the quickest and smallest backup file output...
>
> Thanks again for help\suggestions
>
You didn't specify the Postgres version - that matters, because older
pg_dump versions (before PG 16) do not support compression. Since PG 16
you can use either -Fc or -Fd (instead of the tar format), and it'll
compress the output using gzip.
Alternatively, you can use --compress=method:level (the supported
methods depend on how the packages were built, no idea what platform
you're on etc.). See
https://www.postgresql.org/docs/current/app-pgdump.html
If you're on older version, you should be able to write the dump to
standard output, and compress that way. Something like
pg_dump -Fc | gzip -c > compressed.dump.gz
However, be aware that pg_dump is more an export tool than a backup
suitable for large databases / quick recovery. It won't allow doing PITR
and similar stuff.
regards
--
Tomas Vondra
