5

Background

I work for a large enterprise; we have a number of systems with interfaces to integrate data between them, some internal and some external websites, and some mobile applications which communicate with on site systems.

There are a number of useful tools available for monitoring various pieces of each of these; a few tools which can tackle a large number of pieces, but nothing which will handle everything (I suspect that's impossible outside of a bespoke system given the variety of things to be monitored; from infrastructure availability, resource usage and capacity planning through to user experience/application level.

A better alternative (in our view) would be to have a single top level system which could be fed by various monitors, and could "report" on the information coming from all of them. By report I mean both giving current status dashboard visibility, providing alerts (email / sms / other), and trend diagrams.

Question

Is anyone aware of a high level (aggregate) monitoring solution?

Requirements

  • Receive data from a variety of third party monitors including such data as:
    • Site Availability / Up time
    • Server Resource Usage / Capacity Planning
    • Message Tracking (where is a piece of data in the system-system interfaces)
    • Event logs
    • Network latency
    • Database blocking
    • (many more...)
  • Allow an aggregated view of this data; e.g.
    • By application (is an application healthy, performing poorly, unavailable)
    • By location (e.g. is an application only unavailable in certain locations due to connectivity issues)
    • When was an issue first & last detected (as opposed to seeing 200 individual notifications generated during the total outage period)
    • Allow users & support engineers to customize what they view (e.g. which applications / functional areas / technologies are they interested in)
  • Generate alerts
    • Via email, but other technologies would be a bonus
    • Allow users to customize what they receive (i.e. so they can decide what's useful and what's spam)
  • Allow information to be linked
    • i.e. support tickets associated with the issue
    • Root cause analysis / communications to be attached (e.g. as with https://status.heroku.com/'s read more sections)
    • Link to other related errors (e.g. to show that a network switch going offline is the cause of other apps becoming unavailable).

Additional Information

Currently we use a mix of the following tools for collecting information:

  • What's Up Gold (CPU, Memory, Disk Capacity, Event Collation)
  • Idera (SQL Server blocking, locking & resource monitoring)
  • SCOM (Site Availability - we piggy back on another business unit's instance so haven't made great use of this resource)
  • New Relic (free) (Site Availability, Throughput, Application Performance)
  • Monitor.US (free) (Site Availability)
  • Bespoke .Net, SQL, PowerShell & AutoIt scripts (service availability, data integrity, biztalk port & orchestration status, biztalk suspended messages, message tracking, system responsiveness monitoring)
  • Various pieces included OOTB with our enterprise applications

The above is quite a rough collection of thoughts; hopefully it's enough to communicate the concept of what we're after though.

I'm currently investigating Data Dog (https://www.datadoghq.com/product/), but wanted to see what else may be out there for this.

  • 2
    Not exactly my area of expertise – but are you asking for Nagios? What's missing in its core usually has a complementing addon (or add-in) to provide it. – Izzy Apr 24 '15 at 15:50
  • 1
    I'm sure you can. If you know what the not-so-techy users need, you probably can pre-configure them a "bundle". There are even Android apps available for it. Nagios IMHO is the most used system in this area, so it is well supported. And due to the way it's designed, little technical knowledge is required to add your own "modules" for something that's not (yet) covered otherwise. Sorry I cannot make it an answer; but I'm not using it myself, and cannot give all the details you've asked for. – Izzy Apr 24 '15 at 16:26
  • 1
    Thanks @Izzy; looks like a good solution to me. I'll try to get an instance setup & configured over the next month or two, then can report back here. – JohnLBevan Apr 26 '15 at 15:45
  • 1
    I keep my fingers crossed for you – but am pretty confident :) When Nagios convinced you at least to some degree, could you please answer your own question with your findings? (And yes, I'm aware that would not be before "the next month or two" :) – Izzy Apr 27 '15 at 9:15
  • 1
    John, looks like you're almost ready to answer your own question – even with multiple choices! Looking forward to that – will certainly be a great article. I'm readying my mouse for the upvote it then deserves :) – Izzy Apr 29 '15 at 10:15

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.