Enterprise and Cloud Monitoring

[Nikumba]

Currently have an older version of PRTG doing monitoring with modem for SMS paging etc, however its not been looked after we have over 1500 sensors paused, no one really checks it etc so it needs replacing/sorting.

I don't want to upgrade it in as a) its old b) the underlying hardware is out of support and we think has issues.

So an option would to build a new PRTG server on new hardware, would prefer a virtual solution but PRTG seems to prefer a physical option plus the advantage of a physical box is if the hypervisor stack goes down you get an alert.

So been looking at options, and of course what I want from a monitoring system is different to what hte apps team want.

From my side of infrastructure I want to know status of my VMware stacks, network switches, firewall, routers, SANS etc.

Apps would want to monitor

• Disk space
• Services
• URL's
• Files (being present/absent/content)
• processe i.e. .exe files running
• SQL, ability to run queries on tables and report based on output
• Monitor scheduled tasks, success/fail
• Uptime
• Who's logged into servers
• IIS
• Oracle
• Event logs for specific entries
• URLs - returning specific values to prove they are up

We are also pushing into 365 with Teams, Office, Exchange online so we would want to monitor that side of things as well.

In order to address the lack of ownership of problems, want to be able to create emails based on alerts so if it was something on the App side would create an email to send into our ticket system for them to look at.

Did look at some previous suggestions but those were a year a go or so, things change.

We are a windows house, but we do have a couple of linux boxes for some bits so can use a Linux based option as well.

Any suggestions gratefully received.

Thanks

 

[BongoHunter]

We're moving all our Monitoring to Elastic (on prem) - they're developing the monitoring side of it quite a lot recently, especially in the last year and it ticks a lot of boxes now for us: https://www.elastic.co/blog/whats-n...ty-7-9-0-unified-agent-kpi-overview-dashboard

Free version is pretty good and works on Windows - but lacks RBAC (it's still got auth). There are plugins out there to add other bits of the paid for version functionality (E.G ElastAlert is well maintained and let's you integrate with other tools to send notifications etc.).

It's one of the few on-prem tools we have found that will scale out well enough for the size of out estate (30k+ hosts) - we're also using Prometheus as a collector where we don't have/want to deploy the Elastic agents.

 

[Throrik]

I'm a big fan of ScienceLogic, if only I could persuade the current place to purchase it. Seeing as you have PTRG at the moment, you do know they do a cloud version? Monitoring is one of those that always seems to get left behind/not looked after - one of your best tools at the end of the day.

 

[Nikumba]

We're moving all our Monitoring to Elastic (on prem)

Thanks I will have a look at this.

I'm a big fan of ScienceLogic, if only I could persuade the current place to purchase it. Seeing as you have PTRG at the moment, you do know they do a cloud version? Monitoring is one of those that always seems to get left behind/not looked after - one of your best tools at the end of the day.

We prefer to keep monitoring on-prem rather than cloud based, we have modem used for SMS alerting.

I totally agree monitoring is one of the best tools, and its annoying its in the state it is, but what happens when no one really uses it or keeps it up to date.

Kimbie

 

[BongoHunter]

I totally agree monitoring is one of the best tools, and its annoying its in the state it is, but what happens when no one really uses it or keeps it up to date.

Kimbie

Do you have a change process? We made checking monitoring for something pre change and again post change mandatory a few years ago and saw a massive change in approach from people (even got teams interested in monitoring non prod!).

 

[Throrik]

We prefer to keep monitoring on-prem rather than cloud based, we have modem used for SMS alerting.

Is that the only reason you're looking to keep it on premise? If so fair enough, but if not there are alternatives like PagerDuty which will take over that for you.