Hi Junior, You have the ability to configure the retry option and count when creating the endpoint. Please refer the below sample configuration,
<endpoint xmlns="http://ws.apache.org/ns/synapse"; name="sample"> > <address uri="http://localhost:8080/sample/echo";> > <suspendOnFailure> > <progressionFactor>1.0</progressionFactor> > </suspendOnFailure> > <markForSuspension> > <retriesBeforeSuspension>5</retriesBeforeSuspension> > <retryDelay>600</retryDelay> > </markForSuspension> > <timeout> > <duration>10000</duration> > <responseAction>fault</responseAction> > </timeout> > </address> > </endpoint> The above configuration is enabling the endpoint to be retried 5 times with a timeout. You can create a replay of the messages to the endpoint with more control by following the store forward integration pattern. Please refer to the documentation [1] for more information. [1] https://docs.wso2.com/display/ESB480/Store+and+Forward+Using+JMS+Message+Stores Thanks and Best Regards, On Mon, Oct 8, 2018 at 2:14 AM Júnior <[email protected]> wrote: > Hi, > > I would like to know if there is a way of retrying an endpoint > automatically in WSO2 ESB. > > As I could see one way of doing that is using the failover mediator with > some leaf endpoints. > > I would like to a scenario like this: > > Try the endpoint and in case of a timeout, try it another two times and > after a third try it would lead to a fault sequence. Using the failover > mediator, it will work only if we enable the suspension of the endpoint. > I would like to have it without adding suspension settings. As I could > see, without the suspension settings, the failover endpoint will try the > request forever until the endpoints respond correctly. > > Is there any way to achieve this scenario? > > My first idea, in case there is no OOTB solution, would be: > > 1. create a sequence that would have the call/send mediator. This > sequence, would have a onError sequence, so in case of a timeout, we would > have a condition for a retry; > 2. In case of timeout, the onError would be called, and we would have a > counter, in case the counter is less than 3, we would call the previous > sequence, else, it would fault. > > Would this be a correct approach for this scenario? > > Thanks, > -- > Francisco Ribeiro > *SCEA|SCJP|SCWCD|IBM Certified SOA Associate* > _______________________________________________ > Dev mailing list > [email protected] > http://wso2.org/cgi-bin/mailman/listinfo/dev > -- *Kamidu Sachith Punchihewa *| Senior Software Engineer | *WSO2 Inc*. (m) +94770566749 / +254778872783 | (e) [email protected] | (ln) kamidu <https://www.linkedin.com/in/kamidu/> <https://wso2.com/signature>
_______________________________________________ Dev mailing list [email protected] http://wso2.org/cgi-bin/mailman/listinfo/dev
