Telegraf Setup FAQ for Troubleshooting

This entry is part 5 of 5 in the series Collecting Performance Metrics

I often get emails asking the same time of questions about the setup of Telegraf for SQL Server and Windows while people are trying to set it up.  Hopefully, this blog post will help answer most people’s questions and help them troubleshoot the setup.

First, if you have a problem with data not showing up run Telegraf from the command line directly and make sure you are not getting any errors.  This typically will show you any errors like if it cannot connect to InfluxDB or something similar.

Second, some people have problems with data showing up in dashboards.  Two issues could be the problem here.  One is there is variable that looks for the data source to match a pattern of InfluxDB-## so it is best to name it the same as the blog article unless you are comfortable editing it in Grafana for each dashboard.  Second, use the dashboards in my GitHub because there are two sets of queries for Telegraf, the dashboards in my Github are using the version 2 which Mark Wilkinson (B | T) work on setting up to make more efficient than the version 1 queries, so those dashboards on the Grafana site or somewhere else do not work with my conf file.

Third, if you having problems with the Windows dashboards, I don’t have the windows counters in the SQL Server conf file if you were wanting to see the SQL Servers there, you will need to copy the input part of the Windows conf into the SQL Server conf and restart Telegraf on the SQL Server.

Lastly, the connection string doesn’t work with Windows authentication, I haven’t explored how or if that will work.  A couple of people have asked but I haven’t gotten around to trying to get that to work.  So do use a SQL authentication account like outlined in the setup blog article.

This will be a working document as I get more questions.

Series Navigation<< Guide for Setting up Telegraf to Monitor InfluxDB

Related Posts

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.