In VMware during VM start the existing disk information is used to configure the VMs. So even if a new disk is created using the new template VM continues to use the old disk.
Once the old root disk is marked for destroy force expunge it and sync the new disk into the VM folder before VM start
If HttpNfcLease initialization fails throw appropriate error
Catch throwable in addition to exceptions during file upload task while importing the OVA into vCenter
Cause - This issue can occur in VMware setups if the guest operating system is not specified when we install the virtual machine. In which case, since we map the OS to 'Other OS', VMwareInstallerISO attachment fails.
Added the necessary guest OS mapping for VMware.
Timeout for ssh related test cases is increased based upon
few test run observations where ssh is taking more time during
restart scenarios as an example. We have done few changes to streamline
ssh library. This is in addition to those to keep less ssh script issues
Signed-off-by: Santhosh Edukulla <Santhosh.Edukulla@citrix.com>
Signed-off-by: SrikanteswaraRao Talluri <talluri@apache.org>
The fix contains fixes related to ssh library majorly.
Complete description should be available under bug
description.
Signed-off-by: Santhosh Edukulla <Santhosh.Edukulla@citrix.com>
Signed-off-by: SrikanteswaraRao Talluri <talluri@apache.org>
Added assertElementInList to cloudstackTestCase.
Users can use this new custom addition to add
assertions, thus replacing current multiple assertions
Signed-off-by: Santhosh Edukulla <Santhosh.Edukulla@citrix.com>
There were few unwanted calls as part of test client, did some clean up
Made the test client API uniform to accept both mgmt and dbsvr details
Did some minor bug fixes as well.
Signed-off-by: Santhosh Edukulla <Santhosh.Edukulla@citrix.com>
Conflicts:
tools/marvin/marvin/cloudstackConnection.py
tools/marvin/marvin/cloudstackTestClient.py
tools/marvin/marvin/configGenerator.py
tools/marvin/marvin/deployDataCenter.py
advanced.cfg: Contains three additional flags "useHttps,certCAPath,certPath"
for https usage in marvin for establishing cs connection. We will use the
configuraiton under advanced.cfg provided by user to establish connection over
https. If establishing the connection over https failed, then the default certs
will be used. or else raise the exception, the existing http will work as it
is when useHttps flag set to "False"
Signed-off-by: Santhosh Edukulla <Santhosh.Edukulla@citrix.com>
Signed-off-by: Prasanna Santhanam <tsp@apache.org>
Conflicts:
tools/marvin/marvin/cloudstackConnection.py
The purpose is to verify a given element in list
at a given position with few error checks like
list type,empty list and position
Returns appropriate codes based upon the inputs.
Can be used under tests instead of multiple asserts
Signed-off-by: Santhosh Edukulla <Santhosh.Edukulla@citrix.com>
Conflicts:
tools/marvin/marvin/codes.py
Added Configuration Support to Marvin.
1. It provides the basic configuration facilities to marvin.
2. User can just add configuration files for his tests, deployment
etc, under one config folder before running their tests.
cs/tools/marvin/marvin/config.
They can remove all hard coded values from code and separate
it out as config at this location.
Either add this to the existing setup.cfg as separate section
or add new configuration.
3. This will thus removes hard coded tests and separate
data from tests.
4. This API is provided as an additional facility under
cloudstackTestClient and users can get the
configuration object as similar to apiclient,dbconnection
etc to drive their test.
5. They just add their configuration for a test,
setup etc,at one single place under configuration dir
and use "getConfigParser" API of cloudstackTestClient
It will give them "configObj".They can either pass their own
config file for parsing to "getConfig" or it will use
default config file @ config/setup.cfg.
6. They will then get the dictionary of parsed
configuration and can use it further to drive their tests or
config drive
7. Test features, can drive their setups thus removing hard coded
values. Configuration default file will be under config and as
setup.cfg.
8. Users can use their own configuration file passed to
"getConfig" API,once configObj is returned.
Signed-off-by: Santhosh Edukulla <Santhosh.Edukulla@citrix.com>
Conflicts:
tools/marvin/marvin/cloudstackTestClient.py
tools/marvin/marvin/integration/lib/utils.py
This will add time taken in seconds along with start and end timestamp
logging facility for test cases.Currently this is not
available for console logs
Signed-off-by: Santhosh Edukulla <Santhosh.Edukulla@citrix.com>
Fixed following test cases:
1) test_01_volume_from_snapshot
2) test_03_snapshot_detachedDisk
3) test_07_template_from_snapshot
All three test cases are running successfully on XenServer but test cases (01 and 07) failing on KVM due to a common issue.
Product defect logged for this: https://issues.apache.org/jira/browse/CLOUDSTACK-5097
test_03 is running ok on KVM too, the problem was - disk was not getting detached because it was not unmounted before detaching. Added code to unmount the disk.
Signed-off-by: Girish Shilamkar <girish@clogeny.com>
Changes:
- Set total capacity of a host if it has changed in the CapacityChecker thread
- Fix bug while setting the reserved/used cpu/mem capacity - only one of them used to get set
Fix is use the storage overprovisioning factor (global configuration parameter "storage.overprovisioning.factor") to calculate total provisioning capacity for storage space allocation over VMFS based storage pools as well.
There are two level of thin provisioning provided in VMware, storage level and file-level (VMDK) thin provisioning. in CloudStack, all volumes are provisioned with thin VMDK format, so at hypervisor level, we ALWAYS do thin provisioning. If storage vendor has the ability to provide storage level thin provisioning in addition to VMDK thin provisioning, it is also allowed since it is transparent to Cloudstack.
Signed-off-by: Sateesh Chodapuneedi <sateesh@apache.org>
used cpu is getting bumped up when the over provisioning factor > 1. This was because we didnt record the overprovisioning factors of the vms which got deployed pre 4.2
Upgrade path will fix that by populating the cpu/mem overprovisioning factors for each of the vms in user_vm_details table using the global overprovisioning factor.
Reviewed by : bharat kumar <bharat.kumar@citrix.com>
Signed off by : nitin mehta<nitin.mehta@citrix.com>
-Support passing isHidden option as closure, for conditional hide/show
based on context
-Adds new option format for isHidden option:
return false == show field and all columns (default)
return true == hide only header and form column
return 2 == hide header and form, as well as returned item column -- these
item columns will be skipped over entirely, not just hidden
Likitha Shetty
SrikanteswaraRao Talluri
Frank.Zhang
Wei Zhou
Bharat Kumar
Gaurav Aradhye
Harikrishna Patnala
venkataswamybabu budumuru
Santhosh Edukulla
John Kinsella
Sebastien Goasguen
Girish Shilamkar
Alena Prokharchyk
Prachi Damle
Jessica Wang
Ashutosh K
Edison Su
Sheng Yang
Sateesh Chodapuneedi
Nitin Mehta
Brian Federle