| From: | Josh Berkus <josh(at)agliodbs(dot)com> |
|---|---|
| To: | pgsql-hackers(at)postgresql(dot)org |
| Subject: | Re: Can pessimistic locking be emulated? |
| Date: | 2003-02-27 23:16:40 |
| Message-ID: | 200302271516.40835.josh@agliodbs.com |
| Views: | Whole Thread | Raw Message | Download mbox | Resend email |
| Thread: | |
| Lists: | pgsql-hackers |
Merlin,
Just as a suggestion: In most of my applications, we have a security layer
which is implemented through server-side functions. These functions keep a
table updated which contains:
lock_table
record_id
lock_user
time_locked
This allows us to avoid nasty "your update cannot be processed"-type error
messages by showing the user up front which records are locked, as well as
allowing the admin to decide when locks should "time out".
I tend to find in general that database locking mechanisms are a very poor
locking strategy for a good UI.
--
Josh Berkus
josh(at)agliodbs(dot)com
Aglio Database Solutions
San Francisco
| From | Date | Subject | |
|---|---|---|---|
| Next Message | Justin Clift | 2003-02-27 23:46:56 | Re: analyze after a database restore? |
| Previous Message | Alvaro Herrera | 2003-02-27 22:44:11 | Re: analyze after a database restore? |