Red Hat NETWORK 4.0.5 - CHANNEL MANAGEMENTT GUIDE Guide de dépannage télécharger pdf (Page 36)

F5 SDC Troubleshooting Guide

Proprietary and Confidential Information of F5 Networks

6 Performance

HTTP Performance is Degraded

6.1.1 Error Description

The TPS of the HTTP routing is much slower than expected.

6.1.2 Causes

 Keep-alive: server/client peers are not using keep-alive.

 Number of maximum connections (Max Connections Count Limit (Per

Server) configured when adding a Remote Peer) between the HTTP server peer

and the server is too small to support the traffic load.

 Number of maximum connections between client peer to client is too small to

support the traffic load.

 The HTTP virtual server disconnects after each response.

 Hosting machines are not strong enough. VM is being used.

 Size of messages is too big.

6.1.3 Symptoms

 Timeouts appear for the client or in the CPF. Timeouts by our side appear in the

CPF log with the format “Peer Timeout event occurred for {0}, message {1}”.

 On the server side there are many TCP connections that are in a state:

TIME_WAIT.

6.1.4 Resolution

 Check that the HTTP virtual server’s configuration for Close Connection on

Answer is not enabled.

 Verify that Keep Alive is enabled when configuring an HTTP peer

(server/client).

 Increase the Max Connections Count Limit (Per Server) configuration for the

HTTP server peer.

1 2 ... 31 32 33 34 35 36 37 38 39 40 41 42

Pas de commentaire

Red Hat NETWORK 4.0.5 - CHANNEL MANAGEMENTT GUIDE Guide de dépannage Page 36