previously working BL event gives timeout error (STILL)

General
1

sorry for double post, but original http://support.backendless.com/t/previously-working-good-bl-event-now-gives-timeout-error-again is marked as “Solved” and I guess not actively monitored.
The problem isn’t fixed in any way, I have still timeout errors for my custom events.

2

update: it’s started working properly now.
Still, what was it, how to prevent it, and what to do in order next time this will be fixed quicker?

3

The problem is internal to our implementation. The root cause of it is in synchronization between multiple instances of JS CodeRunner. When a task is executed, it is done through a user account which has special permissions for the external hosts approved for your app and a permission for your file repo. We’re seeing that sometimes the user accounts are not properly replicated between the machines in the cluster and it causes the error you’re seeing. It is a priority item for us and we’re working on a resolution.

Regards,
Mark

4

Thanks Mark! So, even though we called custom events that didn’t used external host, the problem is related to these external hosts, interesting.

5

I don’t know should I report that or you monitoring already, but we again started to get same error

6

now custom events working normal, although queries somewhat slower

7

It should be working now - we have made changes to stabilize the system.

8

For couple of hours now I see return of the same error for my custom events

9

Is it the one which hits an external host?

10

No, it’s more like ALL custom events that I tested (3-4), and these custom events have no external calls.

11

Could you please check if the individual queries the code makes result in the timeouts?

12

well, I have dummy custom event that do almost nothing:

Backendless.enablePromises();
Backendless.ServerCode.customEvent('myCustomEvent5', function(req) {
    console.log('myCustomEvent5, context: ' + JSON.stringify(req.context));
    console.log('args: ' + JSON.stringify(req.args));
    return Promise.resolve({res:'okeedoc!'});
});

and it’s gives same timeout error

13

Ok, seems working now.
Still curious, why it’s breaks every day for last several days. (once again, these are events not calling any external host, nor making big workload)

14

Hi Yuriy,

It should be working now. Could you please check?

Regards,
Mark

15

Yup, seems working now.
Still curious, why it’s breaks every
day for last several days. (once again, these are events not calling any
external host, nor making big workload)

16

The issue is not related to what your app does specifically. It is a side effect of the shared environment with thousands of apps where just a few of them can create undesired consequences. We are working to improve it, but the time is not always on our side. This is one of those examples where for productions apps Backendless Pro or Managed Backendless provide a significant advantage.