function readOnly(count){ }
Sign Up ›
Starting November 20, the site will be set to read-only. On December 4, 2023,
forum discussions will move to the Trailblazer Community.
Learn More

Browse by Topic
ShowAll Questionssorted byDate Posted
+ Start a Discussion
aholetecaholetec 

Outbound Message Queue not being processed fully!

Dear All,

I have a question relating to outbound messages.  It is not directly associated with APEX, however I could not find an appropriate section to log it under. 

The situation is as follows:
1.) We have a data loading process that loads hundreds of thousands of records into an Unlimited org
2.) For this entity we have an outbound message that is processed on the other end by a .NET Web Service listener.

However, we are noticing that the queue just keeps increased, and the old items are not actually consumed by the Web Service Listener.  Is there a cap on the number of outbound messages that can be processed by SFDC at one time?  We have checked the IIS and it seems that it can handle large volume of requests, so we are skeptical that IIS is the cause.

Does the SFDC outbound message queue work by batching up outbound messages, or does it directly execute them as soon as it encounters them?

Any help on this would be great appreciated!

Cheers,
Alex
SuperfellSuperfell
There's an intermediate queue. Are the old rows listed with a failure reason ?
aholetecaholetec

Hi,

Most of the messages in the queue have 0 attempts with no reason and
some of them have "java.net.SocketTimeoutException: Read timed out".

Regards,
Alex
SuperfellSuperfell
the read timeout exception means your listener did not respond within the set timeout (of 10 seconds), or more likely there's some firewall that just silently eating the attempt to send the message to you. Once this happens enough time your endpoint gets put in the bucket with all the other endpoints that are slow to respond, and attempts to deliver your messages will be made less often. Once your endpoint has responded correctly and within a decent performance window, you will be migrated back to the regular delivery attempts schedule.