From: | Bryce Nesbitt <bryce2(at)obviously(dot)com> |
---|---|
To: | Alvaro Herrera <alvherre(at)2ndquadrant(dot)com> |
Cc: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>, pgsql-bugs(at)postgresql(dot)org |
Subject: | Re: BUG #12834: Avoiding Disk Control Writes for better Laptop / SSD compatibility. pgstats |
Date: | 2015-03-06 17:57:56 |
Message-ID: | CAC9LFPd_XJ2kJsww6sZkpFeGFrrcyZ9Dtb0U0Vz1SpkxhpRsqA@mail.gmail.com |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-bugs |
On Fri, Mar 6, 2015 at 6:19 AM, Alvaro Herrera <alvherre(at)2ndquadrant(dot)com>
wrote:
> Use "SHOW autovacuum" in psql to avoid being confounded by
> mind-controlling superpowers.
>
Indeed. My mistake. Autovac *was* the droid I was looking for. On orders
of the Empire, that droid has been eliminated on twenty five laptops.
Millions more exist however, each chewing through the write cycles on SSD's
in a futile attempt to bring down the social order.
--
More seriously: I've been cataloguing what churns the SSD at idle. The top
two are Google Chrome's constant rewriting of HTML5 Local Storage elements,
and PostgreSQL. This is true both in terms of raw writes and blocks
written. It's a significant fraction of the SSD's life (some 40% in the
first laptop analyzed above).
This will never show up in a PostgreSQL benchmark, and never cause a
problem for the PSQL team. But it's eating through SSD's wherever default
mode Postgres is installed but not heavily used.
From | Date | Subject | |
---|---|---|---|
Next Message | Kevin Grittner | 2015-03-06 18:33:58 | Re: BUG #12834: Avoiding Disk Control Writes for better Laptop / SSD compatibility. pgstats |
Previous Message | Alvaro Herrera | 2015-03-06 14:19:34 | Re: BUG #12834: Avoiding Disk Control Writes for better Laptop / SSD compatibility. pgstats |