Monitoring Microsoft Exchange
By Ryan Johnson, Systems Engineering Manager
Brad Bolton, Systems Engineer
Microsoft Exchange is a business-critical application. If it goes down, no one will be able to send email, receive email, or access their calendars…which means that everyone in the company from the clerks to the CEO will be screaming.
To ensure that this does not happen, monitoring Microsoft Exchange is vital. Here are the top five things that we recommend that you monitor in Exchange 2010, 2013 or 2016…
- Where is the Exchange database?
Exchange provides high availability through the use of Database Availability Groups (DAGs), whereby a problem with one server causes the database to automatically failover to another database in the DAG. For many organizations, there’s a good chance that when this happens, the database that will now be in use will be located on a server that is not geographically friendly. For example, although your people might all be located in California, if something goes wrong with the local server things might automatically failover to North Carolina. While this keeps things up and running, the distance can slow things down and degrade the user experience.Out of the box, Exchange does these failovers without creating any alerts that a failover has happened. To keep things running smoothly, though, we have found it is imperative that you know where the mounted copy of the database is at any given time. The key to staying on top of this is to monitor when failovers actually happen. Ultimately, you should always be aware of where your databases are in your DAG. If you know when databases failover, you can move them back or deal with them accordingly.Keep in mind that with Exchange’s DAGs, you don’t have to be in a “major disaster” situation for a database to automatically failover. An automatic failover can also be triggered by simple things such as a brief network outage, degraded performance on the host or storage, and more.Another issue here is that to optimize performance you want to have your databases spread out throughout your machines. You don’t want to let them all end up on one host, while your other machines are sitting idle. Proactive monitoring Microsoft Exchange will alert you as to which databases are out of place.
- Are the Exchange databases and indexes healthy?
For Exchange to work properly, both the databases and their associated indexes must be both mounted and healthy. For monitoring Microsoft Exchange databases, you want to be alerted if a database’s status is not either “mounted” or “healthy.” For monitoring Microsoft Exchange indexes, you want to be alerted if an index’s status is not “healthy.” If a database or index status is anything else, such as “suspended,” “disconnected” or “seeded,” you want to know.
- Are the Exchange services and components functioning properly?
It is vital that you monitor all of Exchange’s core services, as well as all of the dependent services that are involved. In monitoring these things, you need to keep in mind that the fact that services are running does not necessarily mean that they are functional. For Exchange services to be functioning properly, the components need to be functioning properly, too. Unfortunately, monitoring the components is a key area that often gets overlooked.To monitor Exchange services and components, simply watch to see if each item is started or not started. If it’s running, that’s a good sign! But if any service or component gets knocked off line for any reason, a timely alert will let you quickly investigate and resolve the issue.
- Are the DAG copy and replay queues backing up?
Within the DAG, it’s also critical that you monitor the size of the copy and replay queues. If either of these exceeds a stated threshold, you want to know. This is because if there’s a backup in either of these queues, the downstream machines within your DAG will not be current. In the event of a failover, you would be failing over to a database that was not fully up-to-date.What should the backup threshold level be? This depends on your specific environment and requirements. A huge organization with an extremely fast network will probably run on very tight tolerances. A company with a more limited budget that does not have massive bandwidth can expect to see a little backup in the queues during peak times of the day.When the queues do exceed your desired threshold, you’ll want to troubleshoot immediately, before the queue grows even larger. Typically you’ll find that a backup in the copy queue is due to a network-related problem. Aggressive backups in the replay queues are generally a sign that either the storage has been degraded or there’s a performance issue between the storage and the host.
- Are your internal and external mail transport queues backing up?
To keep mail flowing smoothly, another vital aspect of monitoring Microsoft Exchange is to ensure that your internal and external mail transport queues do not exceed a pre-determined threshold. These transport queues can be significant pain points.For example, say someone in the company sends an email out to a list of all 20,000 of your employees worldwide. A few minutes later, thousands of these people start replying. Next thing you know, you have a mail storm and your internal mail transport queues are all plugged up! Or say something else goes wrong, and suddenly Facebook stops accepting mail from your domain. Which means everything heading to Facebook gets stuck in your external transport queues, clogging things up for all of your outbound mail. Proactive monitoring will quickly expose problems like these so that they can be addressed.
How do you get all of this Exchange monitoring in place?
Coyote Creek has built customizable automated monitoring systems that provide a complete “heartbeat” of the Exchange infrastructure, handling all of these monitoring tasks and more. As Microsoft Gold Certified Partners with over 20 years of experience helping enterprise-level organizations make the most of Microsoft technology, we like to think that nobody knows how to monitor this stuff better than we do!
Which is why the easiest way to get all of this Exchange monitoring in place is to give us a call. You don’t have to invest in big box monitoring solutions. By hiring Coyote Creek you can have monitoring solutions that are catered to you and your needs…at a fraction of the cost. We can get things set up so that you can then handle all of the monitoring yourself, or we can provide a full 24/7/365 remote monitoring and management solution for you. Just let us know what you need.