I am actually trying Timescale (latest version on postgres 14 locally installed on my laptop with default settings) for one of my client and i got some issue about disk space.
I am actually inserting backup data’s row by row from a custom multithread C# application.
after few days on intense insert, I am running out of disk space (500Gb) while backup (mainly csv file) file size are < 1Gb in zipped file
When I summarize all chunk_compression_stats() information’s:
size=2817Mo and size_compress=218Mo.
When compression size is interesting, the real database size is terrible.
Do you have any idea why this happen and how to investigate?
Hi @rems welcome to the community!
Do you have any log files that could take up that much space? Also, seems like you have a hypertable with compression enabled, could you share what’s the compression policy? (if any)
I have changed some settings in my application witch seems to improve the disk space.
Before inserting old data (from old application backups) i was executing
SELECT alter_job(job_id, next_start => ‘infinity’) from timescaledb_information.job_stats WHERE hypertable_name LIKE ‘t%_%’
I changed for
SELECT alter_job(job_id, scheduled => false) from timescaledb_information.job_stats WHERE hypertable_name LIKE ‘t%_%’
and it looks better.
I checked on log file and i have 50Gb of log… i will investigate on this. thanks for the help