Skip to main content
Announcements
Qlik Introduces a New Era of Visualization! READ ALL ABOUT IT

How to collect Snapshots for high memory usage in Replicate

No ratings
cancel
Showing results for 
Search instead for 
Did you mean: 
David_Fergen
Former Employee
Former Employee

How to collect Snapshots for high memory usage in Replicate

Last Update:

Sep 3, 2020 10:05:27 AM

Updated By:

David_Fergen

Created date:

Sep 3, 2020 10:05:27 AM

Description:

In order to troubleshoot high memory usage, we need to collect memory snapshots for the Replicate service or the task that's having high memory usage.

Steps:

1) Need to verify which repctl process is causing a high memory. If its a repctl task or the repctl process.

One way to do this in Windows server by going to the Task Manager.

Open Task Manager and go to Details tab

David_Fergen_0-1599141426286.png


Right click on the column names and select columns

David_Fergen_1-1599141426293.png


Scroll down until you see Command Line and check this option and click OK

David_Fergen_2-1599141426298.png


A new column will show up that says Command Line. Look for the repctl.exe under the names tab

David_Fergen_3-1599141426303.png


If you expand the command line, it will show you the tasks that are currently running. If there are 2 tasks running, you will see one repctl process for each task.


2)Then check if this process is related to a specific task or if it is related to a Replicate server.

3)Resume the task (the task that caused too high of memory) , make sure to set all components to verbose. 

4)Also, repsrv.log with Verbose on all the components. 

5)While the memory is increasing run memory reports at least every 5 minute interval. (See explanation below)

 

Explanation on how to troubleshoot memory issue:

 

==> If the memory issue ( is increasing) in the server:

You can run this command with a task’s name that doesn’t exist, and then it will give a memory report for entire replicate server:

 

Repctl.exe -d <data_dir> connect; getsnapshot task=mytask; disconnect

 

==> If the memory issue ( is increasing) in a specific task:

then, get memory snapshot for the task (or diag package)

 

In both cases, if it's a server memory issue or task memory issue:

1) For 15 minutes Open Verbose debug level on all components at the relevant log (server log,task log)

2) Provide memory report 3 times (one report every 5 minutes).

(If it fails fast provide as many memory reports as you can)


Here is how to generate the Replicate memory report:
On Linux, from the bin directory, use repctl to get a snapshot of the memory report.
The task name in this example is ora2ora. Replace it with the actual task name:

. ./arep_login.sh
./repctl connect\; getsnapshot task=ora2ora\; disconnect


You will have two files created (or appended) at the logs directory:

pools_report.txt
pools_statistics.txt

Each time you execute the "getsnapshot" command, these files will be updated.
Note, for non-default data directory, you should use the “-d data_dir” option of the repctl command:
./repctl -d data_dir connect\; getsnapshot task=ora2ora\; disconnect

Please include the repsrv.log + entire task log + Diagnostic packages + memory reports that matches the time frame when you have the memory issue.

Labels (1)
Comments
danielrf
Contributor II
Contributor II

This is an example if the replicate server is linux, you need to add the port and make sure the task is running, otherwise you will get a success but no files.

 

[attunity@server bin]$ /opt/attunity/replicate/bin/repctl.sh -d /local/attunity/data/ connect port=1024\; getsnapshot task=F0911_test\; disconnect
[connect command] Succeeded
[getsnapshot command] Succeeded
[disconnect command] Succeeded
[attunity@server bin]$ find /local -name "*pools_*" -print
/local/attunity/data/logs/pools_statistics.txt
/local/attunity/data/logs/pools_report.txt
/local/attunity/data/logs/F0911_test_pools_report.txt
/local/attunity/data/logs/F0911_test_pools_statistics.txt
[attunity@glsdvlapp365 bin]$ date
Thu May 16 16:30:15 CEST 2024
[attunity@server bin]$ date
Thu May 16 16:36:47 CEST 2024
[attunity@server bin]$ /opt/attunity/replicate/bin/repctl.sh -d /local/attunity/data/ connect port=1024\; getsnapshot task=F0911_test\; disconnect
[connect command] Succeeded
[getsnapshot command] Succeeded
[disconnect command] Succeeded
[attunity@server bin]$ date
Thu May 16 16:42:43 CEST 2024
[attunity@server bin]$ /opt/attunity/replicate/bin/repctl.sh -d /local/attunity/data/ connect port=1024\; getsnapshot task=F0911_test\; disconnect
[connect command] Succeeded
[getsnapshot command] Succeeded
[disconnect command] Succeeded
[attunity@server bin]$ ls -ahl /local/attunity/data/logs/F0911_test_pools_*
---------- 1 attunity attunity 77K May 16 16:29 /local/attunity/data/logs/F0911_test_pools_report.txt
---------- 1 attunity attunity 46K May 16 16:29 /local/attunity/data/logs/F0911_test_pools_statistics.txt

Contributors
Version history
Last update:
‎2020-09-03 10:05 AM
Updated by: