WATS On-premise system monitoring

Updated and applies to WATS Server version 2016.2

Overview

WATS On-premise servers should if possible be monitored to reduce the risk of unexpected downtime. This document serves as guidelines for implementing such monitoring.

Availability monitoring

The web application should answer http 200 OK when accessed. 

• Requesting http://<servername>/ should return http 200 and html login form
• Requesting http://<servername>/wats-dc/reportcenter.svc should return http 200 and a wcf service info page

Performance metrics

Monitoring SQL Server performance:

• Long running requests (>5sec)
• High memory demand
• Persistent high IO / disk activity / Queue length
• Deadlocks

Monitoring IIS/Application server performance
WATS Application normally runs under application pool called WATS. There are a lot of performance counters related to IIS server, but most of the relevant counters can be found under the ASP.NET v4.0.30319 category

Recomended IIS Server monitoring

• High memory usage or extensive disk
• Memory leakage / worker process recycling

Proactive / failure prediction

• Predicted disk space outage