Thread: Pg_bulkload and speed
Hello! Due to speed up loading of a table of about 4 fields with 10 millions of record, are there some parameters to optimize? In same machine it takes about 15 minutes, in other machines about one hour. Parallel mode is better than direct mode? Other interesting things? Thank you in advance for your kind interest! Francesco
On 4/29/15 4:24 AM, Job wrote: > Hello! > > Due to speed up loading of a table of about 4 fields with 10 millions of record, are there some parameters to optimize? > In same machine it takes about 15 minutes, in other machines about one hour. There's a huge number of things that could account for this. Is postgresql.conf the same? Same class of hardware? Same OS version? Same filesystem? Same OS and filesystem configs? How are you actually loading the data? > Parallel mode is better than direct mode? What 'parallel mode'? -- Jim Nasby, Data Architect, Blue Treble Consulting Data in Trouble? Get it in Treble! http://BlueTreble.com
Hello, > Due to speed up loading of a table of about 4 fields with 10 millions of record, are there some parameters to optimize? > In same machine it takes about 15 minutes, in other machines about one hour. > > Parallel mode is better than direct mode? > Other interesting things? Yes, parallel mode is typically faster than direct mode in pg_bulkload. I wonder why there is so large difference in load speed as you noted... Regards, ++++++++++++ Takashi Ohnishi <onishi_takashi_d5@lab.ntt.co.jp>