Thanks for writing back, but I don't think so. There's no message of a deadlock in the log, and the first query started at 12:25, the next one 12:31, 12:39, 12:50, 12:54, so there's plenty of time in between.
On Tue, Apr 1, 2014 at 1:01 PM, Hoover, Jeffrey <jhoover@jcvi.org> wrote:
Could they both be trying to update the same row, resulting in a deadlock?
I'm using postgresql 9.0.13, and I have a simple query that seems to be stuck. I did a
postgres=# select procpid, query_start, waiting, current_query from pg_stat_activity;
procpid | query_start | waiting | current_query
32605 | 2014-04-01 12:39:46.957666-07 | t | UPDATE public.GL_ACCOUNT_ORGANIZATION SET ROLE_TYPE_ID=$1, FROM_DATE=$2, THRU_DATE=$3, POSTED_BALANCE=$4, LAST_UPDATED_STAMP=$5, LAST_UPDATED_TX_STAMP=$6, CREATED_STAMP=$7, CREATED_TX_STAMP=$8 WHERE GL_ACCOUNT
_ID=$9 AND ORGANIZATION_PARTY_ID=$10
32685 | 2014-04-01 12:25:10.378481-07 | t | UPDATE public.GL_ACCOUNT_ORGANIZATION SET ROLE_TYPE_ID=$1, FROM_DATE=$2, THRU_DATE=$3, POSTED_BALANCE=$4, LAST_UPDATED_STAMP=$5, LAST_UPDATED_TX_STAMP=$6, CREATED_STAMP=$7, CREATED_TX_STAMP=$8 WHERE GL_ACCOUNT
_ID=$9 AND ORGANIZATION_PARTY_ID=$10
As you can see this is a pretty simple update query, but it's been running 24 minutes in one thread and nearly 10 minutes in the other thread. What's also strange is it's not trigger a transaction timeout either.
I've already re-started the database recently, and there's not too many threads executing -- just 38 threads total. Is it possible that the table is corrupted or needs repair?
