This morning I received an email notifying that our app has been paused. This may be nothing but understanding the cause of the error would be helpful
Sync to MongoDB has been paused for your application: TaskTracker
This is a test app that only I ever use, so there are no connected users and the timeframe in which it was paused I was not using or logged into the app.
Here’s the error text showing in console
TranslatorFatalError Error Mar 20 3:29:00-04:00 0ms SyncError
Error: recoverable event subscription error encountered: maximum number of streams (1000) reached
Source: Error syncing MongoDB write
And some screenshots of the actual error, and then the activity. As mentioned this was not caused by the app but it appears it’s a server level event - which if so, should not generate an error (I would think) unless it was really an error.
This was a temporary issue affecting some clusters, and has since been resolved.
If you have urgent questions about operational issues with MongoDB Cloud services such as Realm, I recommend you contact Cloud Support directly as they can provide more insight into your specific clusters.
@V_P provided a link to their Realm cluster, so this is expected. Access will be limited to their organisation and MongoDB Production Support employees.
I was more looking for clarification about the cause of the error and what the error messages mean and what the effect was/is. The errors in console are often a little vague so what action should be taken, if any, is unclear. For example, what does this mean?
maximum number of streams (1000) reached
Does that mean someone s DDOS’ing us or does it mean server failed on a backup (for example)? Should I restart my instance? Is my data intact? Wouldn’t this be more clear…
Error: recoverable event subscription error encountered: maximum number of streams (1000) reached. The server temporarily paused a backup. No data was affected and it does not require any further action.
I feel as more developers come on board (Realm), understanding what’s going on in the console will be part of the development process (e.g. clearer error messages).
For example, when that event occurred and there were 400 users running my app, would it have crashed?
Is there a cross reference to error messages and what they mean?
This error was due to a backend configuration limit on the sync translator service for your cluster which has been adjusted. The detail on number of translator streams is from a part of the backend sync service you don’t control, so unfortunately there currently isn’t a lot of context for error messages like this unless the person reviewing the logs is a support engineer.
The impact is that you would need to restart the sync service per the email notification you received (which should have included a link to follow).
I absolutely agree! We have work in progress to improve documentation and messaging, and teams are very aware of the feedback requests from users in the forum and support cases (as well as internal users!).
There is a list of Error Codes in the sync client source, but it is missing a lot of context that would go into a complete developer reference. This also doesn’t cover errors that are specific to the sync service infrastructure (like the “TranslatorFatalError” in the first post of this topic).
I’m not aware of an available detailed reference for Realm Sync service error messages including impact and resolution, but I’ll definitely help socialise as soon as one is available .
FWIW I got this same issue for a few of the Realm apps I manage. While the action that could be taken was clear (restart sync) I had no idea if restarting the sync would just make the error happen again, or if it was something about data trying to be sync’d that was wrong - we have user inputted data in our app.
Learning that it was an issue deeper than what we can control is useful - that we SHOULD click the restart sync button, it’s nothing in our particular app, and the error shouldn’t come back. That’s not always the case, and that’s what was unclear IMO.