| From: | jonathan vanasco <postgres(at)2xlp(dot)com> |
|---|---|
| To: | pgsql-general general <pgsql-general(at)postgresql(dot)org> |
| Cc: | Jeff Janes <jeff(dot)janes(at)gmail(dot)com> |
| Subject: | Re: disk writes within a transaction |
| Date: | 2017-03-01 18:19:09 |
| Message-ID: | 63012C9F-FEEB-4AE5-907F-53F27CB7553D@2xlp.com |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-general |
On Feb 17, 2017, at 4:05 PM, Jeff Janes wrote:
> It will probably be easier to refactor the code than to quantify just how much damage it does.
Thanks for all the info. It looks like this is something worth prioritizing because of the effects on indexes.
We had discussed a fix and pointed it; rewriting the code that causes this is pretty massive, and will require blocking out a resource FT for 2 weeks on rewrites and testing. We don't really have time to spare any of those devs, so time to make product tradeoffs ;(
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Zach Walton | 2017-03-01 19:17:02 | Re: Understanding pg_last_xlog_receive_location |
| Previous Message | Steve Atkins | 2017-03-01 18:03:54 | Re: appropriate column for storing ipv4 address |