Why monitoring virtual F5 BIG-IP instances needs to go hand-in-hand with monitoring the underlying hardware platform
Specific F5 BIG-IP hardware platforms support the running of virtual BIG-IP instances, also called vCMP guests. Main monitoring efforts typically focus on these virtual instances (and rightly so!), as they perform the bulk of the work around application delivery.
But one should not forget to monitor the hosting platform, too.
Unless you understand the resource use of the physical hosting platform, you cannot reliably determine the health of your guest vCMPs. For example:
- CPU use of individual vCMPs could be rather moderate, but the hosting platform CPU use is maxed out, limiting vCMP scalability.
- Network traffic of individual vCMPs could be quite reasonable, but the hosting platform NIC use is at full capacity.
The fact is that virtual BIG-IP instances are limited by resources of the hosting BIG-IP platform and, monitoring one without the other, will not point out potential bottlenecks.
HYCU SCOM Management Pack for F5 BIG-IP is able to capture resource use of both F5 BIG-IP physical hardware platform as well as guest vCMPs. That way, you can easily compare historical CPU or network traffic in a single graph or receive alerts when physical or virtual aspect of the BIG-IP deployment starts running out of resources.