Network Performance Monitoring and Diagnostics NPMD or NPM is a modern term we used to describe network system monitoring (NSM), actually for the advance of the technology and generation of changes, modern NPMD or just simple Network Performance Monitoring or Network Performance Management goes through a lot of changes. Do not be misleading by the modern term, it widely used by company in all size, small, medium to large (I refer to multi country and global scale company with multiple datacenter) use them for monitoring anythings the company possess from infrastructure device, WAN, router, switch, network security devices, server, wireless network or anything with IP that company interest to monitoring for availability and performance utilisation. Typically make use of either device poll base (such as SNMP or WMI), flow back (netflow or equivalent ), packet based technology for monitoring. On top of that, it come with option, addon, another product from the same vendor can work together or integration for additional feature such as cover network traffic analysis (NTA), virtualization environment monitoring, cloud monitoring, digital or end user experience monitoring (that use for web application), application performance monitoring (APM), network device configuration change monitoring, VoIP/QoS monitoring etc.
Gartner recently just published latest market guide in March (revised in May) 2020, had saw that and come in conclusion of few things in mind, since being active in supply, implementing, maintaining it for customer since 2005, few insights to share as below:
- The whole world is migrating to cloud, but like it or not, on premise infrastructure will remain, so it makes strategic sense not make full cloud as the solution, it just yet being CloudOps generation. Solution that cover your existing infrastructure, network, server, application or whatever additional needed feature remain the key and you should not be driving too far to new future forwarding technology that had yet being de facto standard, or better to adopt wait and see approach, since in the technology world change is so fast, better technology is coming in year or two, or alternative vendor will come out with better cost effective solution, so you no needed to spend the premium for that. In another consideration is most of new technology you intent to use, it need supporting technology in place enabling it, so it got lot of hidden cost for adopt a new feature that if you based on nice to have, then each year you need to keep doing the same things, since each year will have a new topic or new technology term invented.
- As modern NPMD solution, your existing solution should meet your operation requirements first, whatever the other in the market or what trends should be in second priority to what matters most for you, your existing infrastructure monitoring, or better term since “infrastructure monitoring” for some people do not cover network, server, application monitoring, so in most of the term I prefer to use NSAM instead, to let people know I refer whole infrastructure cover network device, server system (actually customer is use for monitoring services running inside as well, some customer use it for monitoring endpoint/workstation, for those who interest, it can use for that purpose also), application monitoring (we refer enterprise application, in particular for those mission critical enterprise application that running across multiple server as single application, whether it physical or virtual) .
- The Solution should work for you in such a way that there is no need for you to completely replace what you already own. It should allow you the flexibility to addon advance or additional function as module or plug-in basis to extend the function for what you really want, instead of buy a complete full suite of feature, most of the time 99% of your time, you will not use them at all, but you are force to pay for the premium or “maybe” you will needed it in future, or it is something “nice” to have feature.
- Some of the new feature may or may not be necessary to be integrated, in particular for those for “incident” or “troubleshooting”, it can be invest as standalone, since you will not be “continuous” troubleshooting your network, you can step in when it had the case you need to be “troubleshooting”, then just turn on and collecting the data and make the troubleshooting and post recommendation.
- Saw the new stages of IT operations monitoring maturity, now divided into 5 stages. From stage 1: availability monitoring, stage 2 services monitoring, stage 3 diagnostics, stage 4 business insights, stage 5 self-driving ITOM. In reality most of the enterprises possess stage 1 if just need to know whether they incharge infrastructure is all up and running, or performance utilization and availability status. It will cover most of the real environment use cases. If the device is down, and down for how long, what is the whole year’s availability and utilization for capability planning purpose. Stage 2 services monitoring, for those who had server (it can be physical or virtual), you are intent to running to make sure for example web server services in windows server is running (no stopping), since the server pure uptime is mean nothing for this scenario, we want the web server “Services” is up and running. Stage 3 diagnostics or stage 4 insights, depend on your use case, for example, a customer with e-banking application running across multiple server, what really matters for him or her is that what the application performance and SLA given, which component or services we need to upgrade to boost the performance etc, or what is the time for most of the visitor, we can then running campaign to run promotion to shift people login in different time where in historical record is off peak etc. Sometimes we will “outsourcing” or subscribe for public cloud, under that scenario, the most you want to know is the server, application and services are up, you no need to invest extensively and try to duplicate your cloud service provider infrastructure. If they can give you alerts for services down or send you email alerts, or their real time report, something is just work for some customer this way, rather than attempt to duplicate that under your own expenses.
- We saw customers who possess wisdom, know exactly what they want, they know their very own infrastructure and requirements and know what to buffer, they do not react too much with the new technology and features, their own term, “nice” to have, but not what company “needs”. I think when it comes to reading any product literature you need to first know what you want.
- Do not get me wrong, the Gartner report is good, in fact, few vendors listed inside are our supplier, I use their report heavily for various purposes, you just need to know how to use it for your own case.
Feel free to talk with E-SPIN, E-SPIN being active in the consulting, supply, deploy, implementing, architect, integration/modernization/migration other legacy system to modern system accordingly to customer real requirement, complementing hardware and 3rd party solution that work together, custom and tailor made training program for knowledge transfer and operation competency for your team, provide level first, second or third level maintenance support accordingly to your 8×5/24×7 SLA operation required. We are being supplied for global fortune 500, multi national corporation (MNC), national public listed corporation to federal/state government agencies to medium to large enterprise for various operation and strategic requirements, feel free to reach out and we see how we can assist in your requirements.