{"type":"error","message":"failure to subscribe"}

Hi @Romazes and everyone! Thank you for your patience. The team tried to deploy multiple solutions for this issue and was able to reduce the error reports. However, I have shared that this is still repeating, I will update here as soon as they respond.

1 Like

I am also still seeing this, not always, but still quite frequently.

1 Like

As of March 28, this is still a problem. I’m still receiving intermittent “failure to subscribe” error responses when subscribing to the user channel using legacy API keys.

Could you at least eliminate the 30-second freeze when this occurs?

Is this an authentication failure?

Does this only happen when authenticating with legacy API keys or are people encountering this problem when JWT’s are used?

Did authentication get broken when JWT’s were added?

It seems like this would be a rather straightforward bug to fix.

I definitely get the impression that Coinbase has not allocated sufficient resources to supporting the Advanced API product. The documentation has a lot of deficiencies.

As a software engineer I can’t honestly comprehend how it can take over 3 months to resolve such a simple problem.

Hey everyone, :sunglasses:

I’ve been encountering a persistent issue with subscribing to user channels in our product environment. No matter how many attempts I make, the subscriptions just won’t stick. It’s frustrating because it seems like only one out of every ten attempts succeeds.

Oddly enough, I don’t face the same problem when subscribing to the heartbeats channel. That goes through without a hitch.

This issue is causing a lot of inconvenience for our users. Many of them have reported having to constantly restart the client to get their subscriptions working properly.

Could someone from the tech team provide an update on the progress of resolving this issue? It’s impacting user experience significantly, and it would be great to have some insights into what’s being done to fix it.


Hi @Romazes . This is Lisa from Advanced Trade API team. I would like to inform that we recently deployed a code change that should significantly reduce the disconnections/timeout rates. Please let us know if you are still encountering this problem.

This did get much better, 72 hits in the past 30 days… But more importantly it seems to reconnect much faster now which is great.

Last hit April 6th 2024, 16:07:56.552

Hm, interesting observation :sunglasses: