After the upgrade to 21.2.9 the NetOps Portal console service (caperfcenter_console) will not start.
Starting the service it seems to hang for some time then fails to start.
In the DMService.log can see the following error messages:
These are the same messages seen when upgrading to NetOps Portal 21.2.8, as per this KB article:
although the values in the general table are already set to the recommended values:
| PDF.Full.Max.Concurrent | 1
| PDF.Max.Concurrent | 5
The setting email_jobs_thread_pool_size is set to a value greater than the default of 5 in file /opt/CA/PerformanceCenter/DM/webapps/dm/WEB-INF/dm.properties
Release : 21.2.9
Component : PM Web UI Admininistration/Configuration
The following workaround can be used for release 21.2.9:
Check the file /opt/CA/PerformanceCenter/DM/webapps/dm/WEB-INF/dm.properties and verify if the "email_jobs_thread_pool_size" value is set to a higher value than "PDF.Max.Concurrent" value from general table (mysql netqosportal -unetqos -p -e "select * from general where Attribute = 'PDF.Max.Concurrent'").
email_jobs_thread_pool_size = 40
PDF.Max.Concurrent | 5
If so, do the following and set them to the same value of 5:
1. Edit the file:
email_jobs_thread_pool_size = 5
2. Restart the caperfcenter_devicemanager service:
systemctl stop caperfcenter_devicemanager
systemctl start caperfcenter_devicemanager
3. Start the console service:
systemctl start caperfcenter_console
This has been fixed in NetOps Performance Management 21.2.10
Symptom: Errant setting in the Device Manager configuration with Java upgrade can lead to thread pool exception at start-up time. Excessive values set on customer site for the following email job parameters: email_jobs_thread_pool_size and low_priority_email_jobs_thread_pool_size, can generate error when starting Device Manager.
Resolution: With this fix, the executor initialization now sets the email job parameters when higher than out-of-the-box values, and the device manager starts successfully.
(21.2.10, DE532185, 33072606)