[
https://issues.apache.org/jira/browse/CAMEL-14068?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Andrea Cosentino resolved CAMEL-14068.
--------------------------------------
Resolution: Fixed
> Retry subscribe to Salesforce event when server returns 503 Server too busy
> ---------------------------------------------------------------------------
>
> Key: CAMEL-14068
> URL: https://issues.apache.org/jira/browse/CAMEL-14068
> Project: Camel
> Issue Type: Improvement
> Components: camel-salesforce
> Affects Versions: 2.23.4
> Reporter: Tom Wetjens
> Assignee: Andrea Cosentino
> Priority: Minor
> Fix For: 3.0.0.RC3, 3.0.0
>
> Time Spent: 20m
> Remaining Estimate: 0h
>
> Currently we are using the Salesforce component version 2.23.4.
> Every 3 hours the component needs to resubscribe due to expiring token
> (expected behavior). Sometimes however it fails on:
>
> {code:java}
> org.apache.camel.component.salesforce.api.SalesforceException: Error
> subscribing to /event/CustomerLegalEntity__e:
> 403:denied_by_security_policy:subscribe_deniedorg.apache.camel.component.salesforce.api.SalesforceException:
> Error subscribing to /event/CustomerLegalEntity__e:
> 403:denied_by_security_policy:subscribe_denied at
> org.apache.camel.component.salesforce.internal.streaming.SubscriptionHelper$7.onMessage(SubscriptionHelper.java:406){code}
>
> It turns out (not in the logging) that the server returns additional
> information:
>
> {code:java}
> {ext={sfdc={failureReason=503::Server is too busy. Please try your request
> again later.}}, ...}{code}
>
> As described in the Salesforce documentation here:
> [https://developer.salesforce.com/docs/atlas.en-us.api_streaming.meta/api_streaming/streaming_handling_errors.htm]
> the 503 Server too busy is a documented error that should/can be retried.
> It would greatly add to the resilience of our system if the Salesforce
> component would automatically retry a subscribe if failed due to a temporary
> error. The same backoff mechanism as for the reconnect could apply.
>
> Currently our workaround is to detect a 403 (more information is not
> available, like the failureReason) and restart the whole Camel subsystem.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
