locked
Error: NOAUTH Authentication required RRS feed

  • Question

  • We have a NodeJS service that connects to Azure Cache for a job queue.

    This night we started getting errors in the logs, the redis client fails to connect with the following entries;

    The first entries:

    {"level":"error","message":"Redis connection terminated","timestamp":"2017-03-09T00:03:24.399Z"}
    {"level":"error","message":"Redis error: ReplyError: ERR invalid password","timestamp":"2017-03-09T00:03:24.618Z"}
    {"level":"error","message":"Redis error: ReplyError: Ready check failed: NOAUTH Authentication required.","timestamp":"2017-03-09T00:03:24.618Z"}

    Followed by connection failing:

    {"level":"info","message":"wating for reconnect...","timestamp":"2017-03-09T00:53:25.865Z"}
    {"level":"error","message":"Redis connection terminated","timestamp":"2017-03-09T00:53:29.281Z"}
    {"level":"error","message":"Redis error: ReplyError: ERR invalid password","timestamp":"2017-03-09T00:53:29.499Z"}
    {"level":"error","message":"Redis error: ReplyError: Ready check failed: NOAUTH Authentication required.","timestamp":"2017-03-09T00:53:29.499Z"}

    The same code connects fine for our production and development environments, but it currently fails on what we use for quality assurance.

    Node module redis v2.6.5 is used as the client for connecting.

    Any help here would be appreciated.

    Thursday, March 9, 2017 5:26 PM

Answers

  • I am not a node developer, but here are some things I would test:

    1. Are you able to authenticate to the same Redis instance using the same password that is being rejected in the above errors?  This is to make sure that the password didn't get changed accidentally.
    2. If a connection failure occurs, do you re-create the connection object?  Do you need to re-authenticated after such a connection failure?

    If you send the name of the cache and the details of when you started seeing these issues (UTC or include time zone please) to azurecache@microsoft.com, we can also take a look.


    Jon

    Thursday, March 9, 2017 9:41 PM

All replies

  • I am not a node developer, but here are some things I would test:

    1. Are you able to authenticate to the same Redis instance using the same password that is being rejected in the above errors?  This is to make sure that the password didn't get changed accidentally.
    2. If a connection failure occurs, do you re-create the connection object?  Do you need to re-authenticated after such a connection failure?

    If you send the name of the cache and the details of when you started seeing these issues (UTC or include time zone please) to azurecache@microsoft.com, we can also take a look.


    Jon

    Thursday, March 9, 2017 9:41 PM
  • Yes, I can connect to the Redis Cache with another client by using the same authentication credentials.

    As I mentioned, the service has been running fine until recently it suddenly starting to get authentication issues.

    We use a reconnect strategy for disconnects, and it works fine for dev/prod.

    I will send the name of the cache to the supplied email.

    -Frode

    Thursday, March 9, 2017 10:28 PM