Skip Ribbon Commands
Skip to main content
Navigate Up
Sign In

Quick Launch

Average Rating:

facebook Twitter
Email
Print Bookmark Alert me when this article is updated

Feedback

Node intermittently goes down with "[DOM_10022] The master gateway node for the domain is not available."
Problem Description
The issue is intermittent. No detailed error message in the logs. The strace from Integration process points to a 'kill' command killing the process.
Cause

The strace from pmserver points to kill signal terminating the processes. The user has a pre-session task that is calling a script, which runs command 'kill -9 0'. This causes all processes and parent threads to go down. As a result, pmserver, and node processes go down.

 

From strace:


21387 15:17:17.847812 fcntl(5, F_GETFL <unfinished ...>
1025  15:17:17.847900 <... open resumed> ) = -1 ENOENT (No such file or directory) <0.000293>
991   15:17:17.847935 <... mprotect resumed> ) = 0 <0.000276>
919   15:17:17.848043 <... open resumed> ) = 3 <0.001602>
787   15:17:17.848071 kill(0, SIGKILL <unfinished ...>
32400 15:17:17.848222 read(12,  <unfinished ...>
32332 15:17:17.848253 +++ exited with 0 +++
29911 15:17:17.848265 +++ exited with 0 +++
21387 15:17:17.848278 <... fcntl resumed> ) = 0x802 (flags O_RDWR|O_NONBLOCK) <0.000454>
1025  15:17:17.848348 open("/opt/powerctr/rel961/base/baserel/java/jre/lib/amd64/libpmrelrdr.so", O_RDONLY <unfinished ...>
991   15:17:17.848402 write(9, "\30\232\0\0\0\0009http://www.informatica.co"..., 201 <unfinished ...>
919   15:17:17.848437 read(3,  <unfinished ...>
32400 15:17:17.848461 <... read resumed> 0x1448fa0, 8191) = 158 <0.000224>
21401 15:17:17.968680 <... futex resumed> ) = ? <unavailable>
21387 15:17:17.968703 --- stopped by SIGCHLD ---
991   15:17:17.969194 +++ killed by SIGKILL +++
919   15:17:17.969209 +++ killed by SIGKILL +++
787   15:17:17.969219 +++ killed by SIGKILL +++
782   15:17:17.969228 +++ killed by SIGKILL +++
655   15:17:17.969238 +++ killed by SIGKILL +++
651   15:17:17.969247 +++ killed by SIGKILL +++
644   15:17:17.969257 +++ killed by SIGKILL +++
640   15:17:17.969267 +++ killed by SIGKILL +++
Solution
The ​'kill -9 0' command should be run from and command tasks.
More Information
FR PLAT-21370 has been raised so that Informatica can check for 'kill -9 0' command if it is run from pre/post session tasks. If it is, invalidate the task and not run it.
Applies To
Product: PowerCenter
Problem Type:
User Type:
Project Phase:
Product Version:
Database:
Operating System:
Other Software:

Reference
Attachments
Last Modified Date:1/9/2019 7:42 AMID:532626
People who viewed this also viewed

Feedback

Did this KB document help you?



What can we do to improve this information (2000 or fewer characters)