cant connect to redis cluster

[pdeva]

using latest version of medis on mac Version 2.12.1 (3090)

this command from cli works:

redis-cli -c -h xxx.test.xxx.memorydb.us-east-1.amazonaws.com --tls -p 6379

in medis , i did set the correct config params:

it says 'connected' but nothing works...

[luin]

Hey @pdeva 👋

Thanks for using Medis! Did you use a VPN to connect to MemoryDB from your local machine? There is a documentation for connecting to ElastiCache via a jump server but it may not apply to your setup.

Could you please run "CLUSTER SLOTS" in your redis-cli and send the output here? That will help me in troubleshooting the issue.

[pdeva]

i am indeed using Twingate VPN to connect to Memorydb.
the cli command shown above is executed from my desktop so we know that there is no issues with connecting to memorydb itself.

here is the output of cluster slots

clustercfg.test.drktf1.memorydb.us-east-1.amazonaws.com:6379> CLUSTER SLOTS
1) 1) (integer) 0
   2) (integer) 16383
   3) 1) "test-0001-001.test.drktf1.memorydb.us-east-1.amazonaws.com"
      2) (integer) 6379
      3) "d6340974fe9ecefa8a88cf9cd1a3f072cb4bd1ba"
      4) 1) "ip"
         2) "10.0.74.252"
   4) 1) "test-0001-002.test.drktf1.memorydb.us-east-1.amazonaws.com"
      2) (integer) 6379
      3) "412997fc81ef3bcd927167653a39efe0eedd719b"
      4) 1) "ip"
         2) "10.0.85.202"
clustercfg.test.drktf1.memorydb.us-east-1.amazonaws.com:6379>

[luin]

@pdeva Thanks for the details! I managed to set up a MemoryDB on AWS and actually it worked for me. Here is my setup:

I used https://manual.nssurge.com/others/snell.html but it should pretty similar to Twingate so not sure what was the issue in your setup. Can you connect to test-0001-001.test.drktf1.memorydb.us-east-1.amazonaws.com directly with "Standalone" mode in Medis to see if that works?

[pdeva]

you have 1 node in your setup. mine has 2.

[luin]

@pdeva Thanks for the follow-up! I created another cluster that has 2 shards. I was able to connect to the cluster with Medis:

CleanShot.2023-09-10.at.08.14.46.mp4

Is it possible that Medis was able to connect to the clustercfg node but not the individual node?

[DiegoRBaquero]

I am now experiencing this, when I was previously not. What changed? I scaled up the size of the 2 nodes I have from t4g small to t4g medium. 🤔

[luin]

@DiegoRBaquero Sorry for the inconvenience! Nothing should have been changed recently. If it's possible, could you share with me the connection setting that is causing problems? I would troubleshoot it on my side. [email protected].