End User Troubleshooting Steps for SentryOne Monitoring Services
A matter of fact with software solutions is that errors will happen. They can be frustrating to deal with, but fortunately, the SentryOne Support team is here to help. There are a few basic steps you can take to help speed time to resolution for SentryOne SQL Sentry, which is a fancy way to say, “fix the problem fast.”
Ensure Permissions are Correct
Security is as complicated as your security requirements. Start by fully understanding what your requirements are. After that, make sure you’re clear on the available options in SentryOne by reading up on it here.
Below are two elementary considerations:
- SentryOne desires the monitoring service account to have Windows Administrator privileges on any computer with a monitored Windows Task Scheduler instance to collect system-level performance metrics. While it is possible to configure monitoring without this level of security, it will be a simpler administrative experience if you’re able to provide it.
- SentryOne versions earlier than 19.1.1 require sysadmin privileges on each monitored SQL Server and sysadmin is required to monitor targets on SQL Server version 2008 R2 and earlier. If you keep SentryOne up to date, and you’re on newer versions of SQL Server, this is less of a concern.
Ensure Required Ports are Available
Complete information on required network ports are available here.
Like permissions, SentryOne requires specific ports to be open between the monitoring service host(s) and the monitored target(s).
TCP 1433: SQL Server access (or whichever port is being used if nondefault is being used)
TCP 445: Windows Performance Counter access
TCP 135: WMI access*
*You’ll also need one of the following dynamic ranges to be open:
TCP 49152-65535 (Windows 2008 – Vista)
Custom Range: It’s recommended that you start no lower than port 50000 and allocate no fewer than 255 dynamic ports.
Run the Service Configuration Utility
Included in your SentryOne installation is a utility called Service Configuration. This utility is used to point the monitoring services to a new location should you ever need to relocate your database. This utility can also be used to re-initialize the service, resolving many common issues.
Once you run the utility, you will be prompted for the monitoring service account credentials:
Allow a few moments to pass and verify the issue persists.
View Error Messages in the Client
The most obvious error message(s) will be displayed as a red error banner on the Dashboard.
When this happens, you’ll want to double-click on the error banner to open the full error message. Take a screenshot of each listed error to provide to support.
In the bottom right-hand corner of the SentryOne client is a colored circle. This circle should be green, indicating there are no errors. If you note the circle is red, you can double-click on it to bring up the Client Alerts view.
The Client Alerts view will provide information related to different errors present across the installation. Be sure to provide this information to support should you need to submit a ticket.
Enable Error Logging (Help > Logging > Monitoring Service > Errors)
If you’re still encountering errors, ensure error logging is enabled for the monitoring services. Allow a few minutes to pass in order to generate errors in the log. This information will provide better insight into the root cause of many issues.
Retrieve the Error Log
Once the log has been generated, navigate to the SentryOne installation directory on the server hosting a monitoring service and locate the file called rolling-log.txt. Note you may have several of these files and you may need to retrieve this file from other machines if your environment has multiple sites with errors occurring.
Once you’ve located these files, send them to support along with screenshots of any errors or odd behavior you’ve encountered. This information is vital to accelerating the resolution time for the issues you are experiencing.
You can track your cases and share information with other users by creating an account on our support site.
When Nothing Else Makes Sense: Restart Your Monitoring Service(s)
Since the dawn of desktop computers and small office network devices, we have come to understand at least one universal troubleshooting technique.
We do love our operating systems, but there are many subsystems involved in remotely gathering performance data. Some of those subsystems can be unstable at times. If the view into our session with a performance reporting subsystem has destabilized, some very common issues can be resolved by restarting SentryOne monitoring service(s).
In order to restart a service, you’ll need access to the machine hosting it. If you’re unsure of the location of the monitoring services, expand the Monitoring Services folder within a SentryOne site:
I’ve covered some of the most common speedbumps that can impact your SentryOne experience. Our Support team always love to hear from you, but if you’re the type that enjoys digging into things to solve problems yourself, these tips should help with most anything you’ll run into.
In closing, I’ll highlight the two we started with:
- Get security right
- Get the TCP ports right
These two will all but eliminate difficulties that are related to the environment SentryOne is running in.
Dexter is a Customer Success Engineer with a passion for helping customers navigate through SentryOne. With several years of experience on the Support team, Dexter makes it a priority to resolve any issues customers might be facing. Since moving to Customer Success in November 2019, he has begun to take charge in ensuring customers have the knowledge to leverage SentryOne to resolve SQL Server related issues through Tips and Tricks, alert optimization, and blog posts.