[ https://issues.apache.org/jira/browse/CLOUDSTACK-3809?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Harikrishna Patnala resolved CLOUDSTACK-3809. --------------------------------------------- Resolution: Not A Problem > there is calculation mismatch for Total,used , available capacity in CS for > xen cluster and on xen host > ---------------------------------------------------------------------------------------------------------- > > Key: CLOUDSTACK-3809 > URL: https://issues.apache.org/jira/browse/CLOUDSTACK-3809 > Project: CloudStack > Issue Type: Bug > Security Level: Public(Anyone can view this level - this is the > default.) > Affects Versions: 4.2.0 > Environment: branch 4.2;hypervisor =xen > Reporter: prashant kumar mishra > Assignee: Harikrishna Patnala > Priority: Critical > Fix For: 4.2.0 > > Attachments: Logs_XS_DB.rar, screenshot-1.jpg, screenshot-2.jpg, > VMs_with_scaleup_flag_False.jpg > > > Memory calculation on xen host and CS > ================================ > AVL MEM=available memory > ON CS (in MB) || ON XEN HOST(in MB) > > TOTAL 15440.137 || > 16374 > USED 14208 || > 15924 > > AVL MEM 1232.137 || > 451 > Total MEM difference =1232.137 - 451 = 781.137 > Due to this mismatch allocator chose the host but vm deployment fails at run > time > 1-2013-07-25 14:21:32,101 DEBUG [cloud.vm.VirtualMachineManagerImpl] > (Job-Executor-12:job-24 = [ 6f32f9f4-5185-411f-b75e-3c05cbb475f4 ]) > Deployment found - P0=VM[User|test12], > P0=Dest[Zone(Id)-Pod(Id)-Cluster(Id)-Host(Id)-Storage(Volume(Id|Type-->Pool(Id))] > : Dest[Zone(1)-Pod(1)-Cluster(1)-Host(1)-Storage(Volume(29|ROOT-->Pool(1))] > 2- errorInfo: [HOST_NOT_ENOUGH_FREE_MEMORY, 588251136, 469716992] > Steps to reproduce > ----------------------------- > 1-preapre a CS setup with a xen 6.1 host(advance license) > 2-create a service offering say ram (cpu=500,RAM=3.00 GB) > 4- set cluster.(cpu/memory).allocated.capacity.disablethreshold to one > 5-set enable.dynamic.scale.vm to true > 6-deploy some vms using SO ram , SO small instance ( in my setup i was able > to deploy 4 vm with SO ram +1vms with SO small instance) > 7-if vms are scalable then on host AVL MEM= 425 MB and if vms are not > scalable then AVL MEM = 523 (please check screenshot for more details) > Expected > ---------------- > Deployment should be successful > Actual > -------------- > deployment failed with run time error > My observation > ------------------------- > 1->Memory overhead for vm deployed with 3 GB ram is 129 MB and for vm with > 512 MB ram is 49 MB > 2-> four VM 129*4=516 + For two small vm 49*1=49 +system vm overhead 41(cpvm) > +35(ssvm) +34 (rvm) > 3->Total overhead= 516+98+41+35+34=675 MB which is nearly equal to total MEM > difference > 4-->we should also consider memory overhead and should add it in used > capacity to minimize this gap > DB: > ==== > mysql> select * from op_host_capacity where capacity_type=0\G; > *************************** 1. row *************************** > id: 1 > host_id: 1 > data_center_id: 1 > pod_id: 1 > cluster_id: 1 > used_capacity: 14898167808 > reserved_capacity: 0 > total_capacity: 16190157312 > capacity_type: 0 > capacity_state: Enabled > update_time: 2013-07-25 18:21:47 > created: 2013-07-25 15:40:51 > 1 row in set (0.00 sec) > ERROR: > No query specified -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira