Thread: Connection exceeded alert.
Hi all,
I have been trying to find a steps or solution for past 1 year once we received a connection exceeded alert. Couldn't get any answer from anybody or internet so far.
Today I checked a pgbadger report and matched the timeline and found that, at that time of sudden connection hike (not normal), there were 3 queries running at that time (2 update queries on table a , took 5s each, one insert query on table b which took 2.5mns).
Anybody provide me steps what all I have to do 1. At that time of connection hike 2. If I am been asked to check after sometime as connection dropped to analyse.
Please check maximum connection / Session parameter in postgreSQL.conf file and you can increase maximum session/connection.
Please also check locking on database.
2024년 8월 6일 (화) 오전 9:32, Rajesh Kumar <rajeshkumar.dba09@gmail.com>님이 작성:
Hi all,I have been trying to find a steps or solution for past 1 year once we received a connection exceeded alert. Couldn't get any answer from anybody or internet so far.Today I checked a pgbadger report and matched the timeline and found that, at that time of sudden connection hike (not normal), there were 3 queries running at that time (2 update queries on table a , took 5s each, one insert query on table b which took 2.5mns).Anybody provide me steps what all I have to do 1. At that time of connection hike 2. If I am been asked to check after sometime as connection dropped to analyse.
Muhammad Waqas
Senior Technical Support Engineer - Tech Support Center (Karachi)
Mobile: +92-322-2844150
Email: waqas.m@bitnine.net
1st floor, suit no F-04, COLABS Karachi, 8-C Khayaban-e-Tanzeem,
DHA Phase 5 Tauheed Commercial Area,
Defense V Defense Housing Authority, Karachi, Karachi City, Sindh,
75500 Pakistan.
Hi Rajesh,
To effectively troubleshoot this issue, please share your current PostgreSQL configuration values for shared_buffers, work_mem, and maintenance_work_mem.
Additionally, please provide details about the application generating these queries. Are they triggered by POST or GET requests?
To gain deeper insights into query performance, execute the following commands:
EXPLAIN (ANALYZE, BUFFERS) <your_query>;
SELECT pid, usename, query, state, wait_event_type, wait_event
FROM pg_stat_activity
WHERE wait_event IS NOT NULL;
These commands will help identify performance bottlenecks and resource contention.
On Tue, Aug 6, 2024 at 9:32 AM Rajesh Kumar <rajeshkumar.dba09@gmail.com> wrote:
Hi all,I have been trying to find a steps or solution for past 1 year once we received a connection exceeded alert. Couldn't get any answer from anybody or internet so far.Today I checked a pgbadger report and matched the timeline and found that, at that time of sudden connection hike (not normal), there were 3 queries running at that time (2 update queries on table a , took 5s each, one insert query on table b which took 2.5mns).Anybody provide me steps what all I have to do 1. At that time of connection hike 2. If I am been asked to check after sometime as connection dropped to analyse.