Re: 9.4 regression - Mailing list pgsql-hackers

From Greg Stark
Subject Re: 9.4 regression
Date
Msg-id CAM-w4HNms75Rkbxegu5Aw=2Kcg-Pu5H4LbaqfJeA_fzaLkrVFQ@mail.gmail.com
Whole thread Raw
In response to Re: 9.4 regression  (Thom Brown <thom@linux.com>)
Responses Re: 9.4 regression  (Thom Brown <thom@linux.com>)
List pgsql-hackers
Did you report information about the system affected? What filesystem is it on? If it's ext4 does it have extents
enabled?<br/><br />I think on ext3 or ext4 without extents it won't have any benefit but it shouldn't really be any
slowereither since the libc implementation is very similar to what we used to do.<br /><br />I wouldn't expect the
benefitfrom it to have a huge effect either. The main benefit is actually the reduced fragmentation so it would be
prettydependent on the filesystem it's on anyways. Avoiding stalls when creating new wal files is also nice but
shouldn'thappen often enough n your benchmarks to cause an effect on the average numbers.<br /><br /> 

pgsql-hackers by date:

Previous
From: Michael Paquier
Date:
Subject: Re: Should we remove "not fast" promotion at all?
Next
From: Jon Nelson
Date:
Subject: Re: 9.4 regression