Changes:
- Added a two new deployment planners 'UserDispersingPlanner' and 'UserConcentratedPodPlanner' to the DeploymentPlanners
- Planner can be chosen by setting the global config variable 'vm.allocation.algorithm' to either of the following values:
('random', 'firstfit', 'userdispersing', 'userconcentratedpod')
- By default, the value is 'random'. When the value is 'random', FirstFitPlanner is invoked as before that shuffles the resource lists.
- Now Admin can choose whether the deployment heuristic should be applied starting at cluster or pod level. This can be done by using the
global config variable 'apply.allocation.algorithm.to.pods' which is false by default. Thus by default as earlier, planner starts at clusters directly.
'UserConcentratedPodPlanner' changes:
- Earlier to 3.0, FirstFitPlanner used to reorder the clusters in case this heuristic was chosen.
- Now this is done by a separate planner and is applied only when 'vm.allocation.algorithm' is set to this planner
- It reorders the capacity based clusters/pods such that those pods having more number of Running Vms for the given account are tried first.
- Note that this userconcentration is applied only to pods and clusters. Not to hosts or storagepools within a cluster.
'UserDispersingPlanner' changes:
- 'UserDispersingPlanner' reorders the capacity ordered pods and clusters based on number of 'Running' VMs for the given account in ascending order. Aim is to choose thodes pods/clusters first which have less number of Running VMs for the given account
- Admin can provide weights to capacity and user dispersion so that both parameters get considered in reordering the pods/clusters. This can be done by setting
the global config parameter 'vm.user.dispersion.weight'. Default value is 1. Thus if this planner is chosen, by default, ordering will be done only by number of Running Vms, unless the weight is changed.
- HostAlllocators and StoragePoolAllocators also reorder the hosts and pools by ascending order of number of Running VMS/ Ready Volumes respectively for the given account. Thus try to choose that host or pool within a cluster with less number of VMs for the account.
1.load hosts that in maintenance mode because maintenance is no longer an agent status now
2.don't disconnect agent when entering maintenance mode, again it's no longer an agent status now
move all listxxx interface from HostDao to managers(ResourceManager, SecondaryStorageVmManager etc) with decent name using SearchCriteria2
or direct call SearchCriteria2 on demand
Reviewed-by: Alex
Changes:
- When management server starts, it goes through all the pending work items from op_it_work table and schedules HA work for each. It used to mark each item as done. Instead we should keep the item as pending and let it get marked as Done after the HA work is done.
- Changes in VirtualMachineMgr::advanceStop() :
a) if we find a VM with null hostId, we stop the VM only if it is forced stopped.
b) if VM state transition to Stopping fails,for state Starting and Migrating we try to find the pending work item and then do cleanup the VM. In case state is Stopping we can cleanup directly.
c) We proceed releasing all resources only if state transitioned to 'Stopping'.
- Changes in HA:
a) Depend on VirtualMachineMgr::advanceStop() in case host is not found to do VM cleanup
- When Vm state between mgmt server and agent syncs from starting -> running, mark any pending work item as done.
Conflicts:
server/src/com/cloud/vm/VirtualMachineManagerImpl.java
Changes:
To make sure migration does not attempt to pick a host that has running VMs more than the max guest VM's limit:
- Changed manual migration to call host allocators to return a list of hosts suitable for migration. Host allocators check for the max guest VM limit.
- Earlier we returned hosts with enough capacity but now Host Allocators make other checks along with capacity. So the list of hosts returned are hosts that have enough capacity AND satisfy all other conditions like host tags, max guests limit etc. Or in other words Allocators dont return the hosts that dont satisfy all conditions even if they have capacity.
-Therefore, now we mark the list of hosts returned for manual migration as 'suitable' hosts instead of 'hasenoughCapacity' in the HostResponse.
- HA migration already calls allocators, so no change is needed there.
Changes:
- Adding a new table 'hypervisor_capabilities' that will record capabilities for each hypervisor version. Added db schema changes for this.
- Currently a few capabilities have been added, namely, 'max_guests_limit' and 'security_group_enabled'
- Added a new column 'hypervisor_version' to host table. StartupRouting command now takes in this parameter. It should be set when a host connects.
- If a host's hypervisor version is not present, we find all the capabilities rows for that hypervisor type and use the first record.
- 'max_guests_limit' is the maximum number of running guest Vms that a host can have for the given hypervisor.
- Host Allocators use this limit and skip a host if the number of running VMs on that host exceeds this limit.
status 11326: resolved fixed
Also added more logging to the agent rebalance code.
Conflicts:
server/src/com/cloud/agent/manager/ClusteredAgentManagerImpl.java
frank
Abhinandan Prateek
prachi
Alex Huang
anthony
alena
Nitin
Kelven Yang
frank
keshav