| From: | Greg Stark <gsstark(at)mit(dot)edu> |
|---|---|
| To: | Simon Riggs <simon(at)2ndquadrant(dot)com> |
| Cc: | Nicolas Barbier <nicolas(dot)barbier(at)gmail(dot)com>, Markus Wanner <markus(at)bluegap(dot)ch>, pgsql-hackers(at)postgresql(dot)org |
| Subject: | Re: Deadlock bug |
| Date: | 2010-08-25 14:33:10 |
| Message-ID: | AANLkTim5cWMkEq_CBVNFOdrFrrYwQ7o792MyvDZk36a8@mail.gmail.com |
| Views: | Raw Message | Whole Thread | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
On Wed, Aug 25, 2010 at 3:20 PM, Simon Riggs <simon(at)2ndquadrant(dot)com> wrote:
>> FK constraints can also point to non-PK UNIQUE columns.
>
> You're exactly correct and I now understand Markus' comment. Do you
> think that exact meaning prevents my proposal from being useful?
>
I think it just shows it needs more thought. Do we want the nodelete locks to
prevent updates to any unique keys? Or to specify the specific unique
key that it's concerned with? Can we allow multiple nodelete locks on
different keys?
I'm concerned about the proliferation of special types of locks too.
Having lots of different lock types tends to create more deadlocks
rather than eliminate them so this requires some careful analysis of
the interaction with all the other types of locks.
And most importantly :) I don't like the name "nodelete". Maybe "record
pins"? Or "keep locks"?
--
greg
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Tom Lane | 2010-08-25 14:57:28 | Re: Deadlock bug |
| Previous Message | Nicolas Barbier | 2010-08-25 14:23:41 | Re: Deadlock bug |