Re: [Dev] [IMPORTANT] Greasemonkey Tests for MB 3.0.0 - Fail Tolerance
Hi +1. Since in AMQP there are no publisher acknowledgements we cannot guarantee zero message loss. However with publisher side transaction implementation we would be able to do this. Currently running tests with this. State of things as at now is as follows, *Node kill* - Message loss without publisher acknowledgements. Was able to overcome this with publisher transactions *Graceful shutdown* - Indika is looking into this. Was able to fix message loss issues with this (without publisher transactions). Still there are some exceptions thrown that need to be looked at and message redelivery. Network out (OS shutdown) - Hemika did some tests with network out. We had issues with Hazelcast when network is back on. (There were two coordinators in the cluster at the end). @Hemika, Indika Update if I have missed anything. Regards, Asitha On Sat, May 9, 2015 at 8:35 AM, Hasitha Hiranya hasit...@wso2.com wrote: Hi, I suggest we need to do $Subject before release of MB 3.0.0. As MB is used for reliable messaging High availability and fault tolerance is vital. This is the test scenario: 1. Set up a cluster with three nodes (MB1,MB2.MB3) 2. Set up a failover publisher and a failover subscriber. 3. Graceful shutdown/ kill / network out (OS shutdown) MB1/MB2/MB3 one at a time (connected server). We should see 1. No or less spike in publisher throughput 2. No or less spike in subscriber throughput 3. Zero message loss altogether If we pass this test, MB 3.0.0 is pretty stable and sound. Thanks -- *Hasitha Abeykoon* Senior Software Engineer; WSO2, Inc.; http://wso2.com *cell:* *+94 719363063* *blog: **abeykoon.blogspot.com* http://abeykoon.blogspot.com -- *Asitha Nanayakkara* Software Engineer WSO2, Inc. http://wso2.com/ Mob: + 94 77 85 30 682 ___ Dev mailing list Dev@wso2.org http://wso2.org/cgi-bin/mailman/listinfo/dev
[Dev] [IMPORTANT] Greasemonkey Tests for MB 3.0.0 - Fail Tolerance
Hi, I suggest we need to do $Subject before release of MB 3.0.0. As MB is used for reliable messaging High availability and fault tolerance is vital. This is the test scenario: 1. Set up a cluster with three nodes (MB1,MB2.MB3) 2. Set up a failover publisher and a failover subscriber. 3. Graceful shutdown/ kill / network out (OS shutdown) MB1/MB2/MB3 one at a time (connected server). We should see 1. No or less spike in publisher throughput 2. No or less spike in subscriber throughput 3. Zero message loss altogether If we pass this test, MB 3.0.0 is pretty stable and sound. Thanks -- *Hasitha Abeykoon* Senior Software Engineer; WSO2, Inc.; http://wso2.com *cell:* *+94 719363063* *blog: **abeykoon.blogspot.com* http://abeykoon.blogspot.com ___ Dev mailing list Dev@wso2.org http://wso2.org/cgi-bin/mailman/listinfo/dev
Re: [Dev] [IMPORTANT] Greasemonkey Tests for MB 3.0.0 - Fail Tolerance
+1 On Sat, May 9, 2015 at 8:35 AM, Hasitha Hiranya hasit...@wso2.com wrote: Hi, I suggest we need to do $Subject before release of MB 3.0.0. As MB is used for reliable messaging High availability and fault tolerance is vital. This is the test scenario: 1. Set up a cluster with three nodes (MB1,MB2.MB3) 2. Set up a failover publisher and a failover subscriber. 3. Graceful shutdown/ kill / network out (OS shutdown) MB1/MB2/MB3 one at a time (connected server). We should see 1. No or less spike in publisher throughput 2. No or less spike in subscriber throughput 3. Zero message loss altogether If we pass this test, MB 3.0.0 is pretty stable and sound. Thanks -- *Hasitha Abeykoon* Senior Software Engineer; WSO2, Inc.; http://wso2.com *cell:* *+94 719363063* *blog: **abeykoon.blogspot.com* http://abeykoon.blogspot.com -- Blog: http://srinathsview.blogspot.com twitter:@srinath_perera Site: http://people.apache.org/~hemapani/ Photos: http://www.flickr.com/photos/hemapani/ Phone: 0772360902 ___ Dev mailing list Dev@wso2.org http://wso2.org/cgi-bin/mailman/listinfo/dev
