Diagnostics

The Diagnostics page, available to Explorer Administrators, displays key diagnostics for helping to debug Explorer support issues. 

Use this page if:

• Explorer becomes unresponsive for a number of users
• Shift Log is displaying entries in an unusual way
• Comments are not appearing
• Shift Log Entries Not Updating in the Table Until the Workbook is Closed and Reopened
• Search Box is Missing from the Explorer Ribbon

---

Indexing Diagnostics

In Explorer, indexing is used to improve performance when displaying entries and comments.

Entries are used in Case Management, Shift Log, and A-Plus. 

Comments include commentary on an entry, and also comments in trends, pages, and the Info Zone.

If entries and comments are not showing correctly, this is often because something is happening with the index, such as the index being disabled, or an index rebuild is in progress.

The following key metrics are collected.

Is indexing enabled

If indexing is enabled, other metrics will be displayed. 

If indexing is not turned on, open the Explorer Configuration tool on the server, and on the Indexing tab:

• Check that the Cluster URI is correct. This should be available from your network administrator.
• Select the Enable and build index check box.

The time taken to build an index depends on the amount of data in the system, and full results may take half an hour or more to appear. 

Cluster Status

Cluster status refers to the status of the server on which ElasticSearch is installed, and which hosts the cluster used by Explorer. If indexing is enabled, the status results will be one of the following:

OfflineOrUnreachable	The cluster might be inoperational or not configured. Actions: Check the Cluster URI is correct. Check the ElasticSearch server is turned on and connected to the network. Check that ElasticSearch has been installed correctly, including prerequisites.
Unhealthy	ElasticSearch diagnostics is reporting that there could be errors within the Cluster on the Elastic server. Actions: Try rebuilding the indexes. If the problem persists, an administrator who has access to the ElasticSearch server will need to check the Elastic logs for other possible causes. We recommend checking the ElasticSearch website, which has extensive documentation on troubleshooting cluster issues.
Healthy	The cluster is reachable and the indexes are healthy with no known errors. No action is required. Note: On non-production instances, some data might be missing or inconsistent, as usually happens when a database has been restored but the indexes were not rebuilt. If this is the case, rebuild the indexes.

Index Status

Index status refers to the indexing manager in Explorer which connects to the cluster in order to synchronize the index. If indexing is enabled, the status results will be one of the following:

Inactive	The index manager is not active. This usually means that the Index Manager services has stopped and therefore the indexes will be out of date. Actions: Check for errors in the ISS Logger. Rebuild the indexes.
Connected	The index manager is connected to Elastic Server, however Explorer does not yet know the status of the index. This is usually seen upon installation, until the initial index is built, or if the Elastic service is turned off. Actions: Rebuild the indexes.
Rebuilding	The index is in the process of being built or rebuilt. When a rebuild is triggered, indexing will not start until all P2 Explorer services are restarted. Case Management and Shift Log will not show the complete set of entries until indexing has been completed. Note: If an environment has a large number of entries, full results may not appear for half an hour or more. Actions: Check that P2 services have been restarted. Observe that entries in Explorer will appear to be erratic until the index has finished rebuilding.
Synchronizing	Synchronization is active. The index has been built and is working as expected. The index manager will continue to update the index in the background.  No action is required.
Error	The index manager has encountered an error that cannot be automatically handled. Actions: Check for errors in the ISS Logger. Rebuild the indexes. Restart P2 services.

 

---

Server Diagnostics

Displays free physical memory as a percentage, and disk space free for each drive as a percentage. Minimum required resourcing for P2 Explorer and associated modules is described in detail in the Explorer Installation Guide.

A Note on Virtualization: P2 Server, P2 Explorer, and P2 Sentinel are supported in virtual machine (VM) environments.  In addition to the resources required by the host partition, each VM (guest partition) must be provided with sufficient disk resources, processor resources, and memory for each instance specified in the tables below.

For a high usage environment, the server hardware should mirror the VM requirements, at a minimum. If Sentinel will be processing large amounts of data or running frequently, please scale the hardware appropriately. Simply increasing the CPU, RAM, throughput etc, of the Virtual Machine, on a minimum-spec server would not ensure adequate performance due to the capacity of the hardware itself.

Memory Requirements

The memory required on the P2 Server to successfully run P2 Explorer and associated modules depends on the number of Active Users, which can be found on the Licence Audit page. RAM requirements are as follows:

	<15 Active Users	15-50 Active Users	>50 Active Users
Without ElasticSearch	16 GB+	20 GB+	32 GB+
With ElasticSearch	24 GB+	28 GB+	48 GB+

Note: If ElasticSearch is used, allow a minimum of an additional 8 GB of RAM. For large Shift Log installations, an additional 16 GB+ is recommended.

When installing ElasticSearch, the JVM memory setting refers to the memory that Elasticsearch is reserving for itself, not the total memory that is available to the system. If you are noticing performance issues, then increasing this memory allocation may help.

Disk Space Requirements

Each component has the following initial (installation) and ongoing requirements for disk space for P2 Server, P2 Explorer, and P2 Sentinel combined:

Component	Initial	Ongoing Data Store
Application	50 MB	250 MB
Database	1.5 GB	1 GB per 1000 monitors per year

The items that impact database size include:

• Indexing
• Log files
• Imaging
• Backups
• Tags & Calculations
• Events

---

Service Diagnostics

Service diagnostics shows if each Explorer service is running or not. If a service is not running, it will affect the operation of P2 Explorer and associated modules.  

If any of the services have stopped, then it will show an error message. Otherwise, it will show each service as running.

If a service has stopped, open the Explorer Configuration tool and restart it. 

• If the P2 Server service will not start, check the Service Broker is enabled in the P2 Server database. In the Service Broker section, ensure Broker Enabled is set to True.
• If the Data Broker service will not start, check that all your adaptor dependencies have been installed. These are listed in the Software Prerequisites section of the Explorer Installation Guide.
• If you are trying to restart a service and see that it will not stop, open Task Manager and kill the service.

---

Data Broker Diagnostics

This section checks that the Data Broker is functioning correctly. 

• If the Data Broker is running, it will display the time the Data Broker was started, and the status of each Adaptor Pool.
• If the Data Broker is shut down, an error message will appear.

To get adaptor pools to appear, open trends or pages that use datasources.

Adaptor pools are displayed in the format: Pool - {datasource name} {type} and can have the following statuses:

Healthy	The Adaptor Pool is running normally. No action is required.
Locked	Processing of the last request failed due to either the adaptor reporting that it failed, or an exception escaped the adaptor. This results in the adaptor becoming locked. After a few seconds, Explorer automatically tries to unlock the adaptor pool and reinitialize all the adaptors, in case the connection to the historian/database failed. This attempt to unlock will repeat every 5 seconds (or as otherwise configured). Actions: Check for errors in the ISS Logger. Check that the connection to the datasource is stable.
Permanently Locked	Values in the Queue Manager tables for the datasource have been configured in such a way that has locked the adaptor pools, requiring manual intervention. Queue manager tables are used to prioritize requests differently based on configured criteria, so this state indicates an issue with the configuration. Actions: Fix the values in the Queue Manager table for the datasource. Restart the Data Broker service.

---

Release History

• Diagnostics 4.15.0
  • Initial version