Why are we seeing intermittent 503 Errors ?

We are seeing 503 Errors intermittently for some of our API proxies when connecting with the backend. We notice that backend is up and running fine. The CPU, load or memory usage all seem to be normal. Can you please help on this issue ?

Solved Solved
0 1 1,056
1 ACCEPTED SOLUTION

  1. Enabled tracing and found that we were getting 503 due to connection time out errors (> 3 seconds) with the backend.
  2. Looked in the MP logs and found that connection errors were seen only in about 2 out of 8 MPs.
  3. Checked the direct connectivity (using telnet) with the backed on the 2 MPs where 503 errors were seen. We got the connection time out errors.
  4. However, on the other MPs, we were able to connect to the backend without any errors.
  5. The issues was basically because the IP addresses of the 2 MPs were not white listed in the backend.

Whitelisting the IP addresses of the 2 MPs in the backend, resolved the issue.

View solution in original post

1 REPLY 1

  1. Enabled tracing and found that we were getting 503 due to connection time out errors (> 3 seconds) with the backend.
  2. Looked in the MP logs and found that connection errors were seen only in about 2 out of 8 MPs.
  3. Checked the direct connectivity (using telnet) with the backed on the 2 MPs where 503 errors were seen. We got the connection time out errors.
  4. However, on the other MPs, we were able to connect to the backend without any errors.
  5. The issues was basically because the IP addresses of the 2 MPs were not white listed in the backend.

Whitelisting the IP addresses of the 2 MPs in the backend, resolved the issue.