| From: | "Jonah H(dot) Harris" <jharris(at)tvi(dot)edu> |
|---|---|
| To: | Neil Conway <neilc(at)samurai(dot)com> |
| Cc: | Pailloncy Jean-Gerard <jg(at)rilk(dot)com>, pgsql-hackers <pgsql-hackers(at)postgresql(dot)org> |
| Subject: | Re: Concurrent free-lock |
| Date: | 2005-01-24 23:50:10 |
| Message-ID: | 41F589B2.1050904@tvi.edu |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Neil,
Here is some pretty good info on lock-free structures... I'm pretty sure
I tested their code in a multithreaded high-concurrency environment and
experienced the problems I was discussing.
http://www.cl.cam.ac.uk/Research/SRG/netos/lock-free/
Neil Conway wrote:
>On Mon, 2005-01-24 at 08:35 -0700, Jonah H. Harris wrote:
>
>
>>Lock free data structures are cool... but not really applicable to
>>databases. They have a high maintenance overhead, severe complexity,
>>and will fail when there are many concurrent inserts/deletes to the
>>structure.
>>
>>
>
>Can you elaborate on when they would fail, and why?
>
>It might be worth considering lock-free data structures for certain
>parts of the backend, but I'm skeptical they would be much of a win over
>locking most of the time.
>
>-Neil
>
>
>
>
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Neil Conway | 2005-01-25 00:07:07 | Re: Concurrent free-lock |
| Previous Message | Neil Conway | 2005-01-24 23:23:57 | Re: Concurrent free-lock |