A few months ago, I had published a post titled “How to monitor uptime of your websites?”In that post, I had discussed the types of such monitors, as well as a list of uptime tracking tools. During research for that post, I had installed no less than 6 different server alert services for my sites !
Server Monitoring Tools I have used
In second half of 2020, I used the following uptime tracking services (In alphabetical order)
Typically, it is good to have one set of tracking system, and a second for backup. From that perspective, two different uptime services for servers should suffice. But six was an overkill.
Need to rationalize for efficiency
I always intended to reduce the number of server monitoring services towards end of 2020. During this six month time period, what was not anticipated was the sheer number of downtimes (and corresponding alerts from monitoring services). A KVM and OVZ VPS each from Hostslick was down for a long time. So also, VPS from HostSolutions in recent times. I would get multiple email alerts from the monitors, which would really flood my email box! I discontinued some older services and signed up for new ones during Black Friday 2020. Manually, I had to update each of the 6 uptime services. Some site downtime alerts kept throwing false positives.
There was a stark need to consolidate.
Consolidating the Uptime Monitoring Services
By end of December 2020, the decision was made to use three monitoring sites or services, namely,
Uptime Robot, and
From the original six, Bearmon, Netumo and Zilore will have to go. Of which, Netumo and Zilore have already been cancelled.
During Black Friday 2020 sale. I signed up for a paid subscription for HetrixTools for around 30 US dollars a year. Having used this tool the most (and the longest), it was time to opt for paid subscription. Three is still a lot of monitoring services, but better than having six!
Steps Taken to consolidate monitoring services
As I started optimizing the uptime monitoring tools, a couple of oops’es occurred. The email address that I had set up for receiving alerts from Hetrix Tools was one of them. Because a mis-configuration of settings on my part, no email alerts were received when two of my servers went offline. These were brand-new services, and not set up on only two services- Bearmon and Hetrix tools. Fortunately, Bearmon picked it up, since that service is linked to a different email account.
False Positive or server error?
When I logged on to the Virtualizor server management panel for that provider (IonSwitch) the server panel indeed showed the virtual machine to be offline. Logging on to VNC, and running the
command, showed that the server has been online for nearly 26 days!
It may have been a case of false positive from virtualizor, or a software bug. Long story short, the server was not accessible to the outside world. It is likely that the VPS may very well have been running all this time. It was either an external networking issue or something else was going on. This was a clear signal that merely having multiple server monitoring tools is not adequate. One has to use them efficiently and effectively.
The part of my journey where I used the 6 server uptime tracking services has come to its logical conclusion.The combination of three different monitoring tools, is still a log, but I will stay with them for the first part of this year. By mid-2021, my goal is to go back to the original set of two tools: Hetrix Tools and (either of) Uptime Robot and Freshping.
I have also read up on multiple DIY / Self hosted setups for server uptime monitoring, since publishing the original post. Those I have decided to try for a later time.