[MXS-4439] Maxscale is failing with Resource temporarily unavailable errors Created: 2022-12-09  Updated: 2023-01-31  Resolved: 2022-12-17

Status: Closed
Project: MariaDB MaxScale
Component/s: readwritesplit
Affects Version/s: 6.2.0, 6.2.2, 6.4.4
Fix Version/s: 2.5.24, 6.4.5

Type: Bug Priority: Major
Reporter: Pon Suresh Pandian (Inactive) Assignee: markus makela
Resolution: Fixed Votes: 1
Labels: None
Environment:

Linux AZHOBSPPRDPXYSQL01 5.15.0-1021-azure #26~20.04.1-Ubuntu SMP

Attachments: PNG File CPU usage (%).png     PNG File Memory breakdown.png     PNG File Network traffic (bytes per sec).png     PNG File Processes memory usage.png     Zip Archive maxctrl-report.txt.zip     HTML File syslog_091222    

 Description   

Hi Team,

Maxscale service is failing with following errors.

2022-12-01 01:56:55   error  : (52158524) (split-router-PI3); Failed to write message to worker 3: 11, Resource temporarily unavailable
 
2022-12-01 01:56:56   error  : (52083542) (split-router-PI3); Failed to write message to worker 3: 11, Resource temporarily unavailable
 
2022-12-01 01:56:56   error  : (52092117) (split-router-PI3); Failed to write message to worker 3: 11, Resource temporarily unavailable
 
2022-12-01 01:56:56   error  : (52154432) (split-router); Failed to write message to worker 3: 11, Resource temporarily unavailable
 
2022-12-01 01:56:56   error  : (52143297) (split-router-PI3); Failed to write message to worker 3: 11, Resource temporarily unavailable
 
2022-12-01 01:56:56   error  : (52149226) (split-router); Failed to write message to worker 3: 11, Resource temporarily unavailable
 
2022-12-01 01:56:56   error  : (52154987) (split-router-PI3); Failed to write message to worker 3: 11, Resource temporarily unavailable
 
2022-12-01 01:56:56   error  : (52153572) (split-router); Failed to write message to worker 3: 11, Resource temporarily unavailable
 
2022-12-01 01:56:56   error  : (52145246) (split-router-PI3); Failed to write message to worker 3: 11, Resource temporarily unavailable
 
2022-12-01 01:56:56   error  : (52153684) (split-router-PI3); Failed to write message to worker 3: 11, Resource temporarily unavailable
 
2022-12-01 01:56:56   error  : (52131771) (split-router-PI3a); Failed to write message to worker 3: 11, Resource temporarily unavailable
 
2022-12-01 01:56:56   error  : (52092117) (split-router-PI3); Failed to write message to worker 3: 11, Resource temporarily unavailable
 
2022-12-01 01:56:56   error  : (52149239) (split-router-PI3); Failed to write message to worker 3: 11, Resource temporarily unavailable
 
2022-12-01 01:56:56   error  : (52154370) (split-router-PI3); Failed to write message to worker 3: 11, Resource temporarily unavailable
 
2022-12-01 01:56:56   error  : (52145508) (split-router-PI3); Failed to write message to worker 3: 11, Resource temporarily unavailable
 
2022-12-01 01:56:56   error  : (52154981) (split-router-PI3); Failed to write message to worker 3: 11, Resource temporarily unavailable (subsequent similar messages suppressed for 10000 milliseconds)

We increased the "pipe-max-size" limit upto "1073741824"

SQL01:~# cat /proc/sys/fs/pipe-max-size
 
1073741824

But still getting the same error.. Here I have attached the maxscale config , log file and system logs for your reference please check it..

 Comments   
Comment by markus makela [ 2022-12-13 ]

Does adding skip_name_resolve=true under the [maxscale] section help with how often these happen?

Comment by markus makela [ 2022-12-17 ]

The messages between threads in MaxScale will now never be abandoned. This should fix any temporary problems caused by either large peaks in load or due to some blocking operation (domain name lookup) taking too long.

The SystemD watchdog will still kill the MaxScale process if it isn't responding. This should catch any bugs that might've otherwise manifested as an inability to send messages.

Generated at Thu Feb 08 04:28:39 UTC 2024 using Jira 8.20.16#820016-sha1:9d11dbea5f4be3d4cc21f03a88dd11d8c8687422.