problem with load balancer observe/write requests

Hi,

Our project is trying to use cluster of leshan servers with load balancer. We are using dtls and secure connections. Without balancer everything working fine, but with load balancer we have a lot of

2020-09-01 15:09:07.680 DEBUG 8656 --- [/0.0.0.0:5684#5] o.e.c.scandium.dtls.Connection           : Handshake with [/172.19.1.157:56732] has failed
2020-09-01 15:09:07.680 DEBUG 8656 --- [/0.0.0.0:5684#5] o.e.californium.scandium.DTLSConnector   : Handshake with [/172.19.1.157:56732] failed, report error to deferred 1 messages
2020-09-01 15:09:07.680  WARN 8656 --- [/0.0.0.0:5684#5] s.a.l.c.MessageSenderErrorCallback       : onError of message send callback

org.eclipse.leshan.core.request.exception.SendFailedException: Unable to send request coap://172.19.1.157:56732/3/0/15
        at org.eclipse.leshan.core.californium.CoapAsyncRequestObserver.onSendError(CoapAsyncRequestObserver.java:167) [leshan-core-cf-1.1.0.jar!/:na]
        at org.eclipse.californium.core.coap.Message.setSendError(Message.java:1063) [californium-core-2.3.0.jar!/:na]
        at org.eclipse.californium.core.coap.Request.setSendError(Request.java:1131) [californium-core-2.3.0.jar!/:na]
        at org.eclipse.californium.core.network.CoapEndpoint$SendingCallback.onError(CoapEndpoint.java:1456) [californium-core-2.3.0.jar!/:na]
        at org.eclipse.californium.elements.RawData.onError(RawData.java:303) [element-connector-2.3.0.jar!/:na]
        at org.eclipse.californium.scandium.DTLSConnector$2.handshakeFailed(DTLSConnector.java:451) [scandium-2.3.0.jar!/:na]
        at org.eclipse.californium.scandium.dtls.Handshaker.handshakeFailed(Handshaker.java:1425) [scandium-2.3.0.jar!/:na]
        at org.eclipse.californium.scandium.DTLSConnector.processAlertRecord(DTLSConnector.java:1595) [scandium-2.3.0.jar!/:na]
        at org.eclipse.californium.scandium.DTLSConnector.processRecord(DTLSConnector.java:1330) [scandium-2.3.0.jar!/:na]
        at org.eclipse.californium.scandium.DTLSConnector$12.run(DTLSConnector.java:1195) [scandium-2.3.0.jar!/:na]
        at org.eclipse.californium.elements.util.SerialExecutor$1.run(SerialExecutor.java:276) [element-connector-2.3.0.jar!/:na]
        at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149) [na:1.8.0_252]
        at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624) [na:1.8.0_252]
        at java.lang.Thread.run(Thread.java:748) [na:1.8.0_252]
Caused by: org.eclipse.californium.scandium.dtls.HandshakeException: Received 'fatal alert'
        at org.eclipse.californium.scandium.DTLSConnector.processAlertRecord(DTLSConnector.java:1580) [scandium-2.3.0.jar!/:na]
        ... 6 common frames omitted

I’ve tried sending sync/async requests - seems it doesn’t change anything. Leshan version 1.1.0. Could you please help with possible solutions?

please let me know if more details needed.

[sbernard31]

I’m not sure to really get your question, but the answer should be in : https://github.com/eclipse/leshan/wiki/Using-Leshan-server-in-a-cluster#shared-stores

If not, please try to reformulate your question.

[UnforgivenD]

You mean ResponseCallback from LeshanServer.send ? yes, this callback, thanks for explanation.

other a little bit not related question - if I’m using RedisRegistrationStore in servers, and one is failed, device connected to this server still will be able to communicate with new server, chosen by load balancer?