Hi
I am locking some rows using select for update(nowait=true) inside a
transaction.If server in which code exists crashes,locks will be on
rows.Those should be removed manually.I donot want that.I want to keep a
timeout for that transaction.Could you please helpme out with a solution
for this issu
Thanks alot..i have implemented in same format.You people are awesome.Thank
You once again
On Tue, Apr 8, 2014 at 8:58 PM, Hannes Erven wrote:
> Hi,
>
>
> On 2014-04-08 15:27, Vick Khera wrote:
> [...]
>
> Problem 2: you cannot have a second process skip over locked rows.
>>
>
>
> In fact, you
On Tue, Apr 8, 2014 at 11:28 AM, Hannes Erven wrote:
> On 2014-04-08 15:27, Vick Khera wrote:
> [...]
>>
>> Problem 2: you cannot have a second process skip over locked rows.
>
>
>
> In fact, you can: use "FOR UPDATE NOWAIT" and catch any errors.
>
> e.g. SELECT * FROM jobqueue WHERE id=? FOR UPDA
Hi,
On 2014-04-08 15:27, Vick Khera wrote:
[...]
Problem 2: you cannot have a second process skip over locked rows.
In fact, you can: use "FOR UPDATE NOWAIT" and catch any errors.
e.g. SELECT * FROM jobqueue WHERE id=? FOR UPDATE NOWAIT
-hannes
--
Sent via pgsql-general mailing list (p
Problem 1: how (and why) are you locking rows in table Y?
Problem 2: you cannot have a second process skip over locked rows. It
sounds to me like a job queue... what you want to do is make it work
*really* fast to mark the job as taken and commit that work so the
locks are released. Then go do the
Hi
I have two problems to discuss.an you please guide me how to proceed on
this.
*problem 1:*
I have table X('a' is prmiary key) and table Y('b' is primary key).
Table Y has ''a''(X primary key) as foreign key.I have one job which runs
once in one hour.I want to lock the rows in table Y.but this wi
