Tagged: #aborted_connects #abort
December 21, 2015 at 8:57 pm #13463
We have experienced a sudden increase in ‘Aborted connection’ messages in the sql log over last few days. This has led to application outage in some cases.
Here’s an extract from log:
[Warning] Aborted connection nnnnnnn to db: ‘db_name’ user: ‘user_name’ host : ‘host ip address’ ( Unknown error)
1. Our ‘wait_timeout’ variable in MySQL is set to 28800, which is probably the default.
2. The ‘aborted_connects’ system variable has a value around 92000. It increased by 17000 in last 24 hours.
3. Our ‘Collection Interval’ setting is 20 seconds.
Also it’s been 2 weeks since we upgraded to Monyog 6.25 .
December 22, 2015 at 12:52 pm #35589
Please provide us with the following information:
1. Are you using a direct connection or ssh connection to connect to your server through MONyog?
2. Is the user specified in the above error being used in any other applications other than MONyog?
3. Do you have any CSO (Custom SQL Object) enabled in MONyog?
4. Please send us the MONyog.log file, based on your installation and OS it can be found at:
for RPM: /usr/local/MONyog/MONyog.log
for .tar: /MONyog/MONyog.log
Also, please create a support ticket at email@example.com, and we will continue there.
December 22, 2015 at 5:00 pm #35590
1. direct connections
2. user is only used from the application to connect to DB
3. Pre-existing CSOs are enabled
4. Every few hours , a series of messages repeat
WARNING : Possible SQLite DB corruption ( has existed before the outage inour logs)
Received signal SIGKILL/SIGSTOP…….. — IGNORING SIGNAL
Monyog signaled to syop
Monyog is starting
December 23, 2015 at 7:44 am #35591
We have upgraded our SQLite library in our latest release MONyog-6.51, can you please upgrade and see if the problem persists?
Also, please send us the complete MONyog.log file to “firstname.lastname@example.org”, so that we could look into the issue.
You must be logged in to reply to this topic.