Kepler: Issueshttps://projects.ecoinformatics.org/ecoinfo/https://projects.ecoinformatics.org/ecoinfo/ecoinfo/favicon.ico?14691340362010-03-19T20:45:08ZEcoinformatics Redmine
Redmine Bug #4894 (Resolved): One slave doesn't respond correctly when having concurrent requests.https://projects.ecoinformatics.org/ecoinfo/issues/48942010-03-19T20:45:08Zjianwu jianwujianwu@sdsc.edu
<p>When two running workflow are requesting concurrently to one slave, the executions will be twisted. Usually only one workflow can get correct results. I think it is because multiple client requests for the same slave only get the same object reference. And the attributes of the current SlaveController class created a stateful server. A discussion about it can be found at<br /><a class="external" href="http://www.coderanch.com/t/209597/Distributed-Java/java/Are-RMI-Server-objects-stateful">http://www.coderanch.com/t/209597/Distributed-Java/java/Are-RMI-Server-objects-stateful</a>.</p>
<p>A solution to is to refactor the current SlaveController so that its attributes are in a separate class and SlaveController is only a container.</p> Bug #3073 (Resolved): Heartbeat function for slaveshttps://projects.ecoinformatics.org/ecoinfo/issues/30732008-01-16T23:00:35ZChad Berkleyberkley@nceas.ucsb.edu
<p>When a slave registers itself with the earthgrid registry, there needs to be a way to tell if that slave still exists and if it doesn't, the registry entry needs to be removed. If we don't do this, we are quickly going to end up with a lot of dead registry entries since a lot of the time, kepler will not terminate normally.</p>
<p>There should also be an indication in the gui as to whether the slave is alive or not when the user chooses it.</p> Bug #3072 (Resolved): Security issues with distributed executionhttps://projects.ecoinformatics.org/ecoinfo/issues/30722008-01-16T22:58:31ZChad Berkleyberkley@nceas.ucsb.edu
<p>See bug 3071. basically, the issues are that arbitrary code can be executed on the slaves so the utmost care needs to be take wrt security. kar files need to be signed, workflows should also possibly be signed. the command line and scripting actors might need additional security built into them. This basically breaks the java sandbox.</p> Bug #3070 (Resolved): Need access control GUI for slaveshttps://projects.ecoinformatics.org/ecoinfo/issues/30702008-01-16T22:54:05ZChad Berkleyberkley@nceas.ucsb.edu
<p>Kepler needs a gui that allows a user to set access control rules for the local slave if it is running via the gui.</p> Bug #3069 (Resolved): start slave from guihttps://projects.ecoinformatics.org/ecoinfo/issues/30692008-01-16T22:53:10ZChad Berkleyberkley@nceas.ucsb.edu
<p>A user should be able to start a slave controller from the gui instead of starting it from the command line. The command line interface should not be elimated though.</p> Bug #3068 (Resolved): rmiregistry should be started programmaticallyhttps://projects.ecoinformatics.org/ecoinfo/issues/30682008-01-16T22:51:52ZChad Berkleyberkley@nceas.ucsb.edu
<p>A user wanting to use the distributed system should not have to start the rmiregistry manually. kepler should start this for him/her when it starts up.</p>