Vous êtes confronté à des problèmes de performances réseau en raison de demandes de données. Comment allez-vous résoudre les problèmes et améliorer la capacité ?

Use of network monitoring tools is really important. The latest tools provide capabilities to identify specific congestion points with added capabilities self heal fed with predefined logics to hold the fort while you delve deeper, analyze and plan for long term fix. Firstly, identify if it is an expected month end, quarter end or year end rush, or a specific change like, release of a new application, an organization wide change causing higher than normal usage of the network, etc. All these are short term floods expected to be back to normal soon. Please keep in mind all the use cases, why and who uses your network. It is all about traffic originating from one endpoint and trying to reach another inside your network or outside.

Frist identify the probl, by gathers information about the issue with the help of network monitoring tool,isolate the issue with the help of tool like wire shark,analyze and fix and lastly verifying and monitor. Thanks

If you do not know your traffic patterns, you are already running behind. It is very crucial, to segregate network zones, for example, your backup network is separate from higher priority user traffic. So, now there are two scenarios. One where you need to perform analysis and segregate your network into low latency, moderate latency, and traffic which can live with high latency. Second, if you are already there, narrow down to the segment where utilization is high. With modern AI enabled tools you can identify the specific endpoints which are contributing to the high traffic. You can discuss with stakeholders and span the traffic over separate windows in a day.

Hardware upgrades might not be a good solution, unless you are pushed towards it as the only solution. Your stakeholders should understand that high speed, high priority, fault tolerant, multi-path traffic needs costlier infrastructure. There should be a trade off between cost to performance. You are the expert, don't always ask your stakeholders to spend lots of money, rather have a good discussion to define the NFRs. Based on the agreed figures, provide an estimate for a decision. Also, mention what are the RoIs with the cost and what will be the limitations if an investment is not done.

Exactly, this step should come before even you plan for a hardware upgrade. Also, ensure you make it transparent to your stakeholders and showcase how much you have done and how much is possible in future so that you are out of your buffer for a longer period.

Now we are going beyond our scope. We might collaborate to collectively implement cache, but if the network teams start owning data aspects, there will be friction and conflicts. The capabilities can be provided by the network team, but the ownership of optimization of traffic is upto the solution owners. Network team will time and again sensitize about high usage and potential impact, it is an exclusive decision of the solution owners what is their priority. Collaboration with hosting teams by spreading the content servers, nodes, etc., across geography based on usage pattern inputs from network can help big time. You can also save cost with a strategy of follow the sun, etc.