Hi guys, help me out here, I don’t have a lot of experience with network, but I got involved in a project where this situation is occurring.
In my company there are a few gateways that multiple clients can access simultaneously, and we think some clients may have found a breach where if you send the TCP header first you make the gateway wait for the rest of your message and other clients loose performance or are queued. Since it’s a very low latency system the complaints about the performance loss increased. We are having a hard time proofing it though.
With that situation in mind I’ve got two questions:
1 - how can I verify that? What kind of data should I be looking at? My monitoring system is a bit high level, so I don’t actually see the traffic, but knowing what I am looking for will help me ask the responsible team for evidence. 2 - is there a configuration that can be done to the network board to mitigate that? Or that would be handled better by the OS?
No comments:
Post a Comment