Help Center> Elastic Load Balance> User Guide (Kuala Lumpur Region)> FAQ> Monitoring> Why Is There a Large Number of HTTP 499 Errors?
Updated on 2024-04-19 GMT+08:00

Why Is There a Large Number of HTTP 499 Errors?

When you are seeing the HTTP 499 status code, the client has closed the connection while the server is still processing the request.

The possible causes are as follows:

  • The request timeout may not be long enough for the client to send HTTP requests before a connection is closed. Check the request_time field in the access log to view the total time for processing requests and set an appropriate request timeout.
  • Your load balancer may be overloaded with traffic, causing packet loss due to bandwidth limit. Check the outbound bandwidth usage of your load balancer on the Cloud Eye console. For more information, see Monitoring Metrics.
  • The network that connects the client and your load balancer may be unstable, causing long round-trip delay or packet loss. Check the request_time and tcpinfo_rtt fields in the access log or capture packets to check whether the network is normal.
  • The backend server may take a longer time than the request timeout interval to process requests. Check whether the CPU, memory, and network of the backend server have performance bottlenecks.
  • The client closes the connection before receiving a response from the server due to some unknown reasons. Check whether the client closes the connection before an HTTP request is complete.