Why perform Network Management and trends?

[_|brt.drml|_]

Network management or monitoring? 

Scope of the question:

For my company (with a very large network)  I'm writing a research to develop a Network Monitoring capacity for a part of the network. 

Previous phrase was provided by my superiors and in my opinion contains some issues.

First of all, I think that Network Monitoring is not the correct term. I guess it should be Network Management (Containing, performance, configuration netflow stats and alerting).

• performance: knowing what are the current issues on your network, history tracking, 'on' or 'of' state, mapping of visualizing your network.
• configuration: as simple as it should be: compare changes in the config, who made a change, ios tracking...
• Net flow: packet inspection, QoS issues
• Alerting: from all thresholds and prioritizing, alert me with my 'ticketing' tool.

Secondary, the 'part of the network' is not correct approach, as it creates yet another tool in the company. Shouldn't it be better that you have a one-tool-fits all approach? 

The question: 'several - sorry :-)'

• Do you stil use the on-off approach of a device? 
• Do you have a one-fits-all solution? is it with Cisco alone? (DNA ?)
• What have you learned from your tools?
• Do you exploit your data for improvement processes in the current architecture? 
• Are you already learning network programming (DevNet, Yang, Rest, etc)

Why these questions: Just learning and therefore improving my personal knowledge !

I thank you in advance for answering this as it would consume a bit of your precious time.

Sincerely,

Bart 

[Mark Malone]

hi
im answering this coming from a global enterprise company view , but other users may have different views especially to the programming side if coming from that background or smaller companies where they dont pay for such tools like LA and UIM and utilize them a lot more

Do you still use the on-off approach of a device?
What do you mean by this can you elaborate

Do you have a one-fits-all solution? is it with Cisco alone? (DNA ?)
No many companies will use multiple solutions and vendors depending on setup likes of Cisco Prime and Solarwinds Kiwi say for config backups mgmt , use monitoring tools like CA UIM, LiveAction for SD-WAN IWAN monitoring , other topology tools like intermapper

What have you learned from your tools?
Never trust any 1 tool completely to make a critical decision, always look at the fault from every angle including your knowledge of cli as software can get things wrong especially if bugs are in play

Do you exploit your data for improvement processes in the current architecture?
Yes all the time , constantly driving for more secure environment with the balance of full HA resiliency and all while trying not compromising networks speeds

Are you already learning network programming (DevNet, Yang, Rest, etc)
some to work with NCI and ACI, these models can be useful but if your large ent company you may already have tools that do this for you but its worth knowing some , i would say python definitely

net mgmt is separate from net monitoring , you can have a whole setup like aOOB back to firewalls and control all mgmt protocols through the FW and have an isolated mgmt network in times of outages

[_|brt.drml|_]

Hi Mark,

First of all , thank you.

Also I'm looking at this from a global enterprise perspectif. 

Do you still use the on-off approach of a device?
What do you mean by this can you elaborate --> green on, red = down :-) 'I know…, but for many this is the current approach. 

I also have to concur with the 'one' solution approach is not a good option. As I noticed that the 'results' differ when testing with different vendors. (and I will take this in my conclusion)

Bart.

[Mark Malone]

Do you still use the on-off approach of a device?
What do you mean by this can you elaborate --> green on, red = down :-) 'I know…, but for many this is the current approach.

ah yes definitely if i am in front of the device LED status can tell a lot especially on chassis switches with blades but a lot of the times im not and i am reading the actual SNMP alert that was sent to me through the mgmt network

example of what you may receive , it may not always be a hard down green and red , this is telling me my circuit is being hammered at this time and hit 98% utilization , i removed some details for security but then i would go into history checker and see what was causing this at that time


Message: QOS_INTERFACE_UTILIZATIONIN = 98.64 from source XXXXXXXXXXXXX targeting XXXXXXXXX has crossed the level 5 threshold of +90.0-XXXXXXXXXXXXX_Network_Utilization_Alert-

NAS Profile: Network Device Utilization

Time Received: 01/20/20 02:03:25
Time Arrival: 01/20/20 02:03:27
Assigned To:
Count: 4