After upgrading to vRealize Operations Manager 6.4 Objects show no Collection Status
search cancel

After upgrading to vRealize Operations Manager 6.4 Objects show no Collection Status

book

Article ID: 340767

calendar_today

Updated On:

Products

VMware Aria Suite

Issue/Introduction

Symptoms:
  • After upgrading to vRealize Operations Manager 6.4, some objects appear to have no Collection Status.
  • After upgrading to vRealize Operations Manager 6.4, some dashboards do not display any data.
  • In the /storage/vcops/log/analytics-uuid.log file, you see entries similar to:

    Note: The preceding log excerpts are only examples. Dates, times, and environmental variables may vary depending on your environment.


Environment

VMware vRealize Operations Manager 6.4.x

Cause

This issue is caused by inconsistent data types for certain HDS properties. Data types are defined as numeric only, whereas it can also be string data.
 
This results in NumberFormatException errors when starting adapters.

Resolution

This is a known issue affecting vRealize Operations Manager Clusters that are upgraded to 6.4 and had, at one point, imported data from a 5.8.x deployment.
 
To resolve this issue:
  1. Log in to each vRealize Operations Manager Analytic node as a root user.
  2. Run this command:

 su - postgres -c "/opt/vmware/vpostgres/9.3/bin/psql -d vcopsdb -p 5432 -c \"SELECT property_name, min(period_start_time_dt) as min_start_date, max(period_start_time_dt) as max_start_date, property_value_string, count(property_id) FROM hd_resource_property_history_v WHERE lower(property_name) in ('cpu|effective_limit', 'cpu|limit', 'mem|effective_limit', 'mem|limit', 'mem|host_limit') AND lower(property_value_string) in ('unlimited', '-1.0') group by property_name, property_value_string order by 1,2;\"" 

 

      3. If the above command returns a property_value_string of "Unlimited" the next command should also be ran. If there are no occurrences of "Unlimited" the data is fine, and the second command is not needed.

su - postgres -c "/opt/vmware/vpostgres/9.3/bin/psql -d vcopsdb -p 5432 -c \"UPDATE hd_resource_property_history SET property_value_string = '-1.0' where property_instance_id in (SELECT property_instance_id FROM hd_property_instance_v WHERE property_name IN ('cpu|effective_limit', 'cpu|limit', 'mem|effective_limit', 'mem|limit', 'mem|host_limit')) and lower(property_value_string) = 'unlimited';\""

   4. After completing this on all Analytic nodes within the Cluster, you will need to take the Cluster offline via the Admin UI.

To restart the cluster:
  1. Log into the vRealize Operations Manager Admin UI.
  2. Click Take Offline.
  3. Once the Cluster is fully stopped, click Take Online.

    Note: After the Cluster has come online you may need to wait a few collection cycles for the objects to return to a Collecting State.