after network orchestrator refactor, only network elements providingg
services as defined by network offering, are invloved network design and
imlement phase. So OVS network element need to be enables as
'Connectivity' service provider to make GRE tunnels work. This fix
introduced 'Ovs' provider as Connectivity service provider.
To obtain network read/write statistics, multiply sample duration with the
average of the particular performance metric obtained over the sample period.
(cherry picked from commit 16c3f53793)
Signed-off-by: Animesh Chaturvedi <animesh@apache.org>
To obtain network read/write statistics, multiply sample duration with the
average of the particular performance metric obtained over the sample period.
Calculate and update the size of a backed up snapshot. This snapshot size is in turn used to update the secondary_storage count for an account.
(cherry picked from commit e8ba2d0850)
Signed-off-by: Animesh Chaturvedi <animesh@apache.org>
password from the message string before writing to the log. Made a change to do so.
(cherry picked from commit 6d997bd689)
Signed-off-by: Animesh Chaturvedi <animesh@apache.org>
Added fix for exception and listing. Mentioned details under bug.
Post the fix, simulator works fine.
Signed-off-by: Santhosh Edukulla <Santhosh.Edukulla@citrix.com>
Signed-off-by: Koushik Das <koushik@apache.org>
In case of VMware VM during root volume preparation if we are switching to a new volume, force expunge root disk that was created from the old template.
Because otherwise storage garbage collector will later try to expunge the old disk marked for expunge and fail with 'Cannot delete file' exception
since in VMware the new root vmdk has the same name and is now in use.
(cherry picked from commit 94ea2736f4)
Signed-off-by: Animesh Chaturvedi <animesh@apache.org>
In case of VMware VM during root volume preparation if we are switching to a new volume, force expunge root disk that was created from the old template.
Because otherwise storage garbage collector will later try to expunge the old disk marked for expunge and fail with 'Cannot delete file' exception
since in VMware the new root vmdk has the same name and is now in use.
the management server was restarted. The template.properties file created for the
template has the format field in upper-case. This caused the template service to
not to recognise the format and it removed the entry from the template_store_ref
table in db. Fixed the format field in the templatee.properties.
null pointer exception was getting generated when a VolumeTO object was
serialized to create an answer object. If a local storage is used the uri
field will be null. Added null checks for the same.
We hit these excptions whenever a management server held session that was with the old 5.1 vCenter server
is used to make resource calls to the new 5.5 vCenter.
Validate a vCenter session context before it is being used to make a resource call.
And if the context is invalid then discard the context and retrieve a new one.
During the invalidation of an old context handle the context disconnect better
by catching the appropriate exception and returning a newly created context.
vxlan code. Users can set a physical network to isolation type 'vxlan',
put public traffic on that physical network, and it will still attempt
to use 'vlan' isolation on the KVM hosts. This is going to be an issue
for other isolation types as well, but I'm not familiar with them, so
I'm just fixing vxlan for now.
For a detached volume, don't try to find the associated VM on the hypervisor/peer hypervisor host.
By default create a worker VM to perform snapshot operations.
on hyperv. There were multiple issues here. Upload volume was actually
failing because the post download check for vhd on the cifs share was
unsuccessful. Also the agent code wasn't parsing the volume path correctly.
Fixed it too.
Don't package the OVF and VMDK files into OVA after a template is created from volume.
Since packaging process contains reading and writing from the NFS mount, it doubles the amount of data that needs to be moved around
Instead of injecting object of VolumeOrchestrationService into VmwareResource, we now populate the command object (MigrateVolumeCommand here) with required information. Thus we dont need volume orchestration service to query that information from resource.
Signed-off-by: Sateesh Chodapuneedi <sateesh@apache.org>
is attached a hard disk drive is created on the scsi controller. On detach
the data disk is removed from the drive but the disk drive is left behind.
On reattach the agent was again trying to create a disk drive while it was
already present. Fixed the agent code to look up for disk drive while
attaching and if one is not found then only to create the drive for
attaching a data disk.
The agent was always creating a disk with image format vhdx, but the
cloudstack management server defaults to image format vhd for hyperv.
Updated the agent code to be consistent with what cs expects. All disks
are now created with image format vhd.
When VM is not running, existing code is unable to retrieve associated cluster's Id. Now we will try to get this information using previous host where the VM was running.
Signed-off-by: Sateesh Chodapuneedi <sateesh@apache.org>
DetachISO is succeeding even though detach opeartion is failing as cdrom is locked by VM as it was mounted inside VM.
Detect if cdrom is locked or not. If locked fail detach operation and warn user to unmount before detaching the iso/cdrom device.
Signed-off-by: Sateesh Chodapuneedi <sateesh@apache.org>
deployment fails because of that as cloudstack tries to deploy it on a host which is
ctually down. An investigator wasn't present for hyper-v; so cloudstack wasn't able to
determine the status of the host. Wrote an investigator for hyper-v which checks with
other hosts in the cluster for the status of the host being investigated.
host is put in maintenance mode. The migrate flag wasn't set to true in
the maintain answer. This caused cloudstack to not to schedule a migration
work item for vms on the host. Made a change to set the migrate flag to
true in migrate answer.
During VM deployment when base template is being cloned to create VM ROOT disk, get the disk path
i.e. base file name of the VM's ROOT disk from vCenter
vmware.reserve.cpu/mem are set to true. Insted of setting
the ovecommit values to one on upgrade, we popultate them
from the global values.
Signed off by : nitin mehta<nitin.mehta@citrix.com>
brought back up after being down for few hours,snapshot jobs do not get
triggered with reason "there is other active snapshot tasks on the
instance to which the volume is attached".
Murali Reddy
Min Chen
Likitha Shetty
Marcus Sorensen
Hugo Trippaers
Jayapal
Devdeep Singh
Santhosh Edukulla
Anthony Xu
Kelven Yang
Animesh Chaturvedi
Mike Tutkowski
Saksham Srivastava
Anshul Gangwar
Edison Su
Sanjay Tripathi
Rajesh Battala
Sateesh Chodapuneedi
root
Bharat Kumar