From: | Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> |
---|---|
To: | Thomas Kellerer <spam_eater(at)gmx(dot)net> |
Cc: | pgsql-general(at)postgresql(dot)org |
Subject: | Re: Detecting uncommitted changes |
Date: | 2011-12-27 19:22:54 |
Message-ID: | 13168.1325013774@sss.pgh.pa.us |
Views: | Raw Message | Whole Thread | Download mbox | Resend email |
Thread: | |
Lists: | pgsql-general |
Thomas Kellerer <spam_eater(at)gmx(dot)net> writes:
> Bill Moran wrote on 27.12.2011 19:37:
>>> is there a way I can detect if the current session has any uncommitted changes?
>> If I'm understanding you correctly, you could just make it check the
>> transaction status. If there's an active transaction, then there are
>> "uncommitted changes".
> Sounds like what I want, but how do I check the "transaction status" (I'm using JDBC)
More specifically, look to see if the current transaction has assigned
itself a transaction ID. I think the easiest place to see this is in
pg_locks --- it will be holding exclusive lock on a TransactionId object
if so.
There are various cases where you could get a false positive from this
type of test, eg if a subtransaction made some changes and then rolled
back, you'll have an XID even though there's not really anything to
commit. But it will never give a false negative.
regards, tom lane
From | Date | Subject | |
---|---|---|---|
Next Message | Jacques Lamothe | 2011-12-27 19:31:43 | Re: UNABLE TO CONNECT REMOTELY TO port 5436 - CRITICAL |
Previous Message | Raymond O'Donnell | 2011-12-27 19:21:05 | Re: UNABLE TO CONNECT REMOTELY TO port 5436 - CRITICAL |