Hi, With the following changes, I can now see groups with 2 and 3 levels start to spin up VMs:
· As Raj also noted, the group definition seems to have changed. Instead of the “min” and “max” properties I was expecting, it seems I needed to set “groupMinInstances” and “groupMaxInstances” properties. · The format of the runtime information now seems to include a groupInstances level. · The Startos startup sequence changed such that out start_stratos.sh needed to look for a new marker. With that, my goal of getting the new Stratos object model integrated into our code seems to be met, and VMs are now spinning up. The next steps is to verify the right number/type of VMs is coming up (and can be stopped). Thanks all, Shaheed From: Rajkumar Rajaratnam [mailto:[email protected]] Sent: Monday, March 23, 2015 1:37 AM To: [email protected] Subject: Re: Group applications VMs do not start Hi Shaheed, I have found couple of issues in your application definition. You have given the value "0" to groupMinInstances and groupMaxInstances properties in all the group references . So No groups being created and execution stops at that point. "name":"blade", "alias":"blade", "groupMinInstances":0, "groupMaxInstances":0 Please give a positive numbers to these properties and try again. And we don't need to specify the application status ("status":"Deployed") in application definition. It will be set to "Deployed" by auto-scaler once an application is successfully deployed. PS - We should modify our addApplication() rest API in a way that the bean class should not have application status attribute. Otherwise if someone sets this attribute in application definition, it will cause issues in AS. Thanks. On Sun, Mar 22, 2015 at 10:37 PM, Shaheedur Haque (shahhaqu) <[email protected]<mailto:[email protected]>> wrote: As Raj suggested, I moved forward one commit (I see there may be other breaking issues later which are being worked) to: commit 7cb908e3cf4ded92afc46a1792b5bf5dfac1d9b2 Author: R-Rajkumar <[email protected]<mailto:[email protected]>> Date: Sat Mar 21 06:52:24 2015 +0530 fixing some issues with n-level-group-nesting and making a sample commit 31a157dbf7d341a86781b8b40289205f93599333 Author: R-Rajkumar <[email protected]<mailto:[email protected]>> Date: Fri Mar 20 19:51:06 2015 +0530 fixing an NPE in AS And now, the scenario of the first “stuck point” in my email seems to have gone, and now all my group start attempts hit the other “stuck point”: TID: [0] [STRATOS] [2015-03-22 16:45:15,445] INFO {org.apache.stratos.autoscaler.monitor.MonitorFactory} - ClusterMonitor created: ClusterMonitor [clusterId=group3.cartridge-proxy-1x0.cartridge-proxy.domain, hasPrimary=false ] TID: [0] [STRATOS] [2015-03-22 16:45:15,446] INFO {org.apache.stratos.autoscaler.monitor.component.ParentComponentMonitor} - Monitor started successfully: [type] cluster [component] group3.cartridge-proxy-1x0.cartridge-proxy.domain [dependents] none [startup-time] 0 seconds TID: [0] [STRATOS] [2015-03-22 16:45:15,447] INFO {org.apache.stratos.autoscaler.monitor.MonitorFactory} - ClusterMonitor created: ClusterMonitor [clusterId=group3.cisco-sample-vm-1x0.cisco-sample-vm.domain, hasPrimary=false ] TID: [0] [STRATOS] [2015-03-22 16:45:15,448] INFO {org.apache.stratos.autoscaler.monitor.component.ParentComponentMonitor} - Monitor started successfully: [type] cluster [component] group3.cisco-sample-vm-1x0.cisco-sample-vm.domain [dependents] none [startup-time] 0 seconds I’m sure I am doing something silly, but I cannot tell what. Advice welcome. (In the meantime, I’ll pick up the changes Imesh is working on). Thanks, Shaheed From: Shaheedur Haque (shahhaqu) Sent: Sunday, March 22, 2015 3:59 PM To: [email protected]<mailto:[email protected]> Subject: RE: Group applications VMs do not start Thanks, I’ll give that a spin! From: Rajkumar Rajaratnam [mailto:[email protected]] Sent: Sunday, March 22, 2015 1:00 AM To: [email protected]<mailto:[email protected]> Subject: Re: Group applications VMs do not start Hi Shaheed, Please find my comments inline. On Sun, Mar 22, 2015 at 5:58 AM, Shaheedur Haque (shahhaqu) <[email protected]<mailto:[email protected]>> wrote: Hi, Thanks to some recent fixes, I can now setup and deploy, undeploy and remove 2 and 3 tier group applications (I’ll reply to those threads confirming this once I get a bit more testing done). However, none of the VMs referenced ever start. I’m on: commit 31a157dbf7d341a86781b8b40289205f93599333 Author: R-Rajkumar <[email protected]<mailto:[email protected]>> Date: Fri Mar 20 19:51:06 2015 +0530 fixing an NPE in AS You are at one commit before to an important fix. In some cases, the last entry in the log looks like this (and the number of applicationInstances, clusterInstances and members is 0): TID: [0] [STRATOS] [2015-03-21 23:52:31,477] INFO {org.apache.stratos.manager.components.ApplicationSignUpHandler} - Application signup added successfully: [application-id] group3 [tenant-id] -1234 TID: [0] [STRATOS] [2015-03-21 23:52:31,477] INFO {org.apache.stratos.autoscaler.services.impl.AutoscalerServiceImpl} - Application signup added successfully: [application-id] group3 TID: [0] [STRATOS] [2015-03-21 23:52:31,486] INFO {org.apache.stratos.autoscaler.services.impl.AutoscalerServiceImpl} - The application clusters are not yet created. Waiting for them to be created TID: [0] [STRATOS] [2015-03-21 23:52:31,487] INFO {org.apache.stratos.rest.endpoint.api.StratosApiV41Utils} - Application deployed successfully: [application-id] group3 I have found the reason for this issue and fixed it in the following commit. Please take a pull and try again. commit 7cb908e3cf4ded92afc46a1792b5bf5dfac1d9b2 Author: R-Rajkumar <[email protected]<mailto:[email protected]>> Date: Sat Mar 21 06:52:24 2015 +0530 fixing some issues with n-level-group-nesting and making a sample And in other cases, the last entry in the log looks like this (and the number of applicationInstances is 1, but clusterInstances and members are both zero): TID: [0] [STRATOS] [2015-03-21 23:42:07,667] INFO {org.apache.stratos.autoscaler.monitor.MonitorFactory} - ClusterMonitor created: ClusterMonitor [clusterId=group2.cartridge-proxy-1x0.cartridge-proxy.domain, hasPrimary=false ] TID: [0] [STRATOS] [2015-03-21 23:42:07,667] INFO {org.apache.stratos.autoscaler.monitor.component.ParentComponentMonitor} - Monitor started successfully: [type] cluster [component] group2.cartridge-proxy-1x0.cartridge-proxy.domain [dependents] group2.cisco-sample-vm-1x0.cisco-sample-vm.domain [startup-time] 0 seconds TID: [0] [STRATOS] [2015-03-21 23:42:07,675] INFO {org.apache.stratos.autoscaler.monitor.MonitorFactory} - ClusterMonitor created: ClusterMonitor [clusterId=group2.cisco-sample-vm-0x0.cisco-sample-vm.domain, hasPrimary=false ] TID: [0] [STRATOS] [2015-03-21 23:42:07,675] INFO {org.apache.stratos.autoscaler.monitor.component.ParentComponentMonitor} - Monitor started successfully: [type] cluster [component] group2.cisco-sample-vm-0x0.cisco-sample-vm.domain [dependents] none [startup-time] 0 seconds I guess the most likely explanation is that I am still not constructing the Cartridge Group or Application quite correctly. I’ve included below a complete pair as shown by the Stratos CLI. Note that there are quite a few values which are zero or false…could any of them be the issue? I haven't come across this issue. Will look into it. stratos> describe-cartridge-group group3 Service Group : group3 {"name":"group3","groups":[{"name":"blade","groups":[{"name":"slice","cartridges":["cartridge-proxy","cisco-sample-vm"],"groupScalingEnabled":false,"dependencies":{"startupOrders":["cartridge.cartridge-proxy-2x0,cartridge.cisco-sample-vm-2x0"],"terminationBehaviour":"terminate-none"}}],"cartridges":["cisco-sample-vm","cartridge-proxy"],"groupScalingEnabled":false,"dependencies":{"terminationBehaviour":"terminate-none"}}],"groupScalingEnabled":false,"dependencies":{"terminationBehaviour":"terminate-none"}} stratos> describe-application group3 Application: group3 {"applicationId":"group3","multiTenant":false,"alias":"group3","status":"Deployed","components":{"groups":[{"name":"group3","alias":"group3","groupMinInstances":0,"groupMaxInstances":0,"isGroupScalingEnabled":false,"groups":[{"name":"blade","alias":"blade","groupMinInstances":0,"groupMaxInstances":0,"isGroupScalingEnabled":false,"cartridges":[{"type":"cisco-sample-vm","cartridgeMin":1,"cartridgeMax":10000000,"subscribableInfo":{"alias":"cisco-sample-vm-1x0","deploymentPolicy":"static-1","autoscalingPolicy":"economyPolicy","maxMembers":0,"minMembers":0,"artifactRepository":{"alias":"cisco-sample-vm-1x0","privateRepo":true,"repoUrl":"http://xxx:10080/git/default.git","repoUsername":"user","repoPassword":"c-policy-2"}}},{"type":"cartridge-proxy","cartridgeMin":1,"cartridgeMax":10000000,"subscribableInfo":{"alias":"cartridge-proxy-1x0","deploymentPolicy":"static-1","autoscalingPolicy":"economyPolicy","maxMembers":0,"minMembers":0,"artifactRepository":{"alias":"cartridge-proxy-1x0","privateRepo":true,"repoUrl":"http://xxx:10080/git/default.git","repoUsername":"user","repoPassword":"c-policy-2"}}}],"groups":[{"name":"slice","alias":"slice","groupMinInstances":0,"groupMaxInstances":0,"isGroupScalingEnabled":false,"cartridges":[{"type":"cartridge-proxy","cartridgeMin":1,"cartridgeMax":10000000,"subscribableInfo":{"alias":"cartridge-proxy-2x0","deploymentPolicy":"static-1","autoscalingPolicy":"economyPolicy","maxMembers":0,"minMembers":0,"artifactRepository":{"alias":"cartridge-proxy-2x0","privateRepo":true,"repoUrl":"http://xxx:10080/git/default.git","repoUsername":"user","repoPassword":"c-policy-1a"}}},{"type":"cisco-sample-vm","cartridgeMin":1,"cartridgeMax":10000000,"subscribableInfo":{"alias":"cisco-sample-vm-2x0","deploymentPolicy":"static-1","autoscalingPolicy":"economyPolicy","maxMembers":0,"minMembers":0,"artifactRepository":{"alias":"cisco-sample-vm-2x0","privateRepo":true,"repoUrl":"http://xxx:10080/git/default.git","repoUsername":"user","repoPassword":"c-policy-1b"}}}]}]}]}]}} Am I missing some settings? Any advice is most welcome… Thanks, Shaheed P.S. Also, note how the autoscalingPolicy is economyPolicy? Curiously, though I have the logs showing it was added: stratos> list-autoscaling-policies Error in listing autoscaling policies No autoscaling policies found If this is not a related issue, I’ll start another thread for it. I guess this is not a related issue. Thanks. -- Rajkumar Rajaratnam Committer & PMC Member, Apache Stratos Software Engineer, WSO2 Mobile : +94777568639<tel:%2B94777568639> Blog : rajkumarr.com<http://rajkumarr.com> -- Rajkumar Rajaratnam Committer & PMC Member, Apache Stratos Software Engineer, WSO2 Mobile : +94777568639 Blog : rajkumarr.com<http://rajkumarr.com>
