This article outlines the steps on how to fix the error
Search is currently unavailable.
Please contact your Telescope Administrator when searching in TSWeb. OnDemand customers should raise a support ticket for further assistance with this issue.
The error occurs due to memory leaks by other processes on the system leading to a low memory condition on the instance Solr is running on, which can lead to Solr crashing rendering Search unavailable. Most notably, for instances hosted on AWS, earlier versions of the AWS PV driver(7.x) were known to consume a disproportionate amount of memory on the system, as well as LiteAgent, which is another process installed by AWS.
- You can get the environment details of the customer from the article Telescope On-Demand Customer Environment Details.
- Write access to PLNP.
Steps To Fix
Agents should perform the following actions on the App instance of the OnDemand customer.
Log in to the server where the NPS Jetty-Service (Solr) is installed.
Open the Services panel.
Locate the NPS Jetty-Service. Right-click on it and select the Start option on the context menu:
For a permanent fix to this issue, if you're hosted on EC2 instances in AWS you should upgrade AWS PV Driver on the instance Solr is running on to the latest version to fix this issue.
Agents should get a downtime window for OnDemand customers for the above activity and create a ticket on PLNP JIRA of type SaaS Request and priority Medium mentioning the downtime window details. PLNP-5916 is an example of a similar activity.
Additionally, a script can be configured for OnDemand customers which can monitor the status of the Solr service at intervals and restart it if it finds the service not running. PLNP-5903 is an example of a ticket where a script was configured, so agents should mention the possibility of configuring the same if it is possible in the ticket that will be created above.
Once Solr is restarted, Telescope Search should available again.
Please sign in to leave a comment.