Hanging out in IT Operations, we're regularly asked to test the features of Corvil's software on our own networks. For a long time, the capabilities that Corvil's clients could see in their own environments, (electronic trading, mainly), weren't easily transported across to the Enterprise IT. Microburst, RTT values, PNQM, are all very valuable and nice to know, but for 99% of what we do in the Enterprise, they seemed a bit extraneous.
This has changed with the 9.x release of Corvil's product. We're now actively using it on our own LAN and WAN and it's quickly spotting our mistakes. It has already saved us a considerable amount of investigation time and helped us to resolve network performance in a fraction of the time it would've previously taken. Straight out of the box, it was showing us the network traffic that we'd expect to see. After a few minutes’ configuration, I had our WAN connections monitored and displaying on our IT Operations screen in the office. It was quickly apparent what was normal network behaviour and what wasn't.
One example of an oversight on our part was a virtual machine that had been configured in a remote site with the machine itself running in the local site. We quickly spotted that NFS traffic on the WAN was considerably larger than normal. The fix was simple enough to implement, just migrate the machine back to the correct VMFS store. We’d have eventually spotted this mistake, but our Corvil appliance immediately highlighted it and we were able to resolve before anyone else had realized.
Like most companies (surely all companies?) we have nightly backups—these aren't getting any smaller. First thing in the morning, I can see the graphs of the previous night's traffic on the WAN, indicating when the backups were running and when they complete. Sometimes, for various reasons, backups don’t always run as they should. The first item I spot every morning as I walk in the door is our IT Operations screen showing me network utilization. Over-utilization of the link in the early part of the day usually means one of the backups has gone haywire. With that knowledge so easily accessible, it’s something we can quickly resolve.
We introduced caching on our WAN link for all http traffic and we were quickly able to see the impact it had on the WAN. This http traffic, once the top talker with about 90% of the WAN usage, moved to 2nd place with about 37% of the WAN.
After a rollout of Windows 10 to a large number of computers, using the Corvil Appliances, we spotted that traffic to various Microsoft partners had suddenly increased (although not by a huge amount). What was happening here? We eventually figured out that Windows 10 automatically downloaded and installed a number of Windows Store programs without the end user’s knowledge. We were able to adjust our Windows policies accordingly and block this traffic. Clearly not a world-ending issue, but a detail that would be missed without this level of visibility into the network.
Corvil uses Corvil to help ensure your experience is the best it can be. Say that three times fast, or check out a demo here to see what Corvil can do for you.