Ticket: 00838529

Hello, I suggest to add more information into the alerts related to "Virtual host client error". (and maybe other types, but for now I work on this)

For example this real one: "More than 10% of the requests from site fr4-fra to service S:193.58.194.186 failed due to client error."

Me as a user have no aditional info and can only assume what the root cause is. I guess this message is result of some "logs agregation" in this case probably result of having quite some 4xx responses, 4xx is for client error as you know so that could be candidate for such alert. You collect those into the log server and then compare valid requests vs invalid, right?

I dont know what are the other options which might trigger such alarm (like client not finishing 3way handshake, client not sending valid HTTP request which cannot be parsed, ..) but XC could write the reason into the alert. That means, if XC can agregate the logs and create alert based of few conditions, put the condition information into the alert itself, so we customers know why the alert was created. If there are more different client errors happening, then each type should have own alert...

That's my 2 cents

BTW: this was discussed with "Bartlomiej Tarkowski" from F5 XC who understands this topic well.

thx

Zdenek