Link to home
Start Free TrialLog in
Avatar of patron
patronFlag for India

asked on

VMotion & SVMotion in vSphere 4.x and 5.x

Looking for more info regarding vmotion and svmotion approach changed in 4.x and 5.x

•      Any changes done with respect to process execution behind vmotion and svmotion in 4.x and 5.x ?
•       Any port difference  OR mechanism change while we are doing svmotion to shared storage or local/different storage?
•      When doing svmotion..What exactly start  at very first time Mirroring or Disk Copy and how each process get synchronized during svmotion ?
•      VMkernel is always used for all kind of vmotion & svmotion doing online Or Offline?
•      Use of VAAI in storage vmotion ?
Avatar of gheist
gheist
Flag of Belgium image

v5 allows storage vmotion of machines with snapshots
v5 allows vmotion using multiple network adapters

there should not be firewall on vmotion network. that extra millisecond just asks for problems.

generally it makes snapshot-20% , transfers all disk data -60% then it plays back snapshot and deletes original.

Yes

NO, never, that is well documented.
Avatar of patron

ASKER

would need more info to understand  on svmotion backend process changes in 5.x

Role of CBT /Mirror Driver/Datamover if any?

When we start vmotion and svmotion .how exactly it is processed further in 5.x
SOLUTION
Avatar of gheist
gheist
Flag of Belgium image

Link to home
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
Start Free Trial
Avatar of patron

ASKER

will take some time to upgarede it to 6.but its good to see if there is change with respect to svmotion in 6 as well ?
SOLUTION
Link to home
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
Start Free Trial
Avatar of patron

ASKER

Thanks for sharing this for 6, would be great if you please share the same for  DRS chnages made in 5..like how actually it start vomotion/svomtion in 5.x

when we start migration machine or datsore..what actually start and how a process synch with other ongoing processes@backened.

any port /mechanism diffrence when we migrate machine to shared datsore or to single/not shared datastore ?

and was this same in 4.1 ?

for offline vmotion/svmotion it works with managemnet port only ?
SOLUTION
Link to home
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
Start Free Trial
ASKER CERTIFIED SOLUTION
Avatar of Andrew Hancock (VMware vExpert PRO / EE Fellow/British Beekeeper)
Andrew Hancock (VMware vExpert PRO / EE Fellow/British Beekeeper)
Flag of United Kingdom of Great Britain and Northern Ireland image

Link to home
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
Start Free Trial
Avatar of patron

ASKER

Thanks Andrew.

let me explore it more..After upgrading my infra from 4.1 to 5.1..i have observed one common issue ie.
Svmotion times out b/w 26-32 %

i tried to look more into vmkernel logs and found some error through which i checked one vmware KB to disable VAAI
http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=1033665

and strange after this i found svmotion of 700 gb  took max  3 minutes.?

so my concern here is what is difference if i disable values like ...
DataMover.HardwareAcceleratedMove
DataMover.HardwareAcceleratedInit
VMFS3.HardwareAcceleratedLocking

Will there be any difference with svmotion mechanism in both cases like before and after i made changes to values given above.

Please help me to understand, so that we can get this fixed at broad level of infra i have?
Can we do it @SAN side or if Changes required @host level?
I would discuss this changes with your Storage Vendor and VMware Support, if you are suffering Storage vMotion timeouts, something is not correct with your setup or SAN - it's uncommon.
Avatar of patron

ASKER

Thanks a lot Andrew, now I have the solution,but will escalate this to SAN Vendor.
Avatar of patron

ASKER

this was the error i was getting in vmkernel logs earlier before doing changes..

2015-05-26T16:11:08.116Z cpu41:5919261)NMP: nmp_ThrottleLogForDevice:2346: Cmd 0x83 (0x4124c41f9c80, 4751704) to dev "naa.60a98000316b6171412b4459382d4e4a" on path "vmhba5:C0:T15:L57" Failed: H:0x2 D:0x2 P:0x0 Possible sense data: 0xa 0xd 0x2. Act:EVAL
2015-05-26T16:11:08.128Z cpu41:5919261)NMP: nmp_ThrottleLogForDevice:2346: Cmd 0x83 (0x4124c28b4f00, 4751704) to dev "naa.60a98000316b6171412b4459382d4e4a" on path "vmhba5:C0:T15:L57" Failed: H:0x2 D:0x2 P:0x0 Possible sense data: 0xa 0xd 0x2. Act:EVAL
2015-05-26T16:11:08.143Z cpu41:5919261)NMP: nmp_ThrottleLogForDevice:2346: Cmd 0x83 (0x4124c01d8e40, 4751704) to dev "naa.60a98000316b6171412b4459382d4e4a" on path "vmhba5:C0:T15:L57" Failed: H:0x2 D:0x2 P:0x0 Possible sense data: 0xa 0xd 0x2. Act:EVAL


and naa.60a98000316b6171412b4459382d4e4a was the destination Data store..tried for multiple luns but always got above logs for diff destination luns ?

Please advise if this is something we can dig more into this ?
Avatar of patron

ASKER

2015-05-26T16:26:10.911Z cpu42:4050518)WARNING: NMP: nmp_DeviceRequestFastDeviceProbe:237:NMP device "naa.60a98000316b6171412b4459382d4e4a" state in doubt; requested fast path state update...
2015-05-26T16:26:10.911Z cpu42:4050518)NMP: nmp_ThrottleLogForDevice:2346: Cmd 0x83 (0x4124c01fc780, 4751704) to dev "naa.60a98000316b6171412b4459382d4e4a" on path "vmhba2:C0:T16:L57" Failed: H:0x2 D:0x2 P:0x0 Possible sense data: 0xa 0xd 0x2. Act:EVAL
Avatar of patron

ASKER

and then finally got timeout..like....

2015-05-26T16:26:11.242Z cpu33:5163363)ScsiDeviceIO: 3045: Command 0x28 (CmdSN 0xb09a6, World 4751704) to device naa.60a98000316b6164792b457375544948 timed out: expiry time occurs 4ms in the past
2015-05-26T16:26:11.242Z cpu33:5163363)ScsiDeviceIO: 2358: Cmd(0x4124c4252ac0) 0xfe, CmdSN 0xd26 from world 4751704 to dev "naa.60a98000316b6171412b4459382d4e4a" failed H:0x5 D:0x0 P:0x0 Possible sense data: 0xa 0xd 0x2.
2015-05-26T16:26:11.242Z cpu33:5163363)ScsiDeviceIO: 3045: Command 0x28 (CmdSN 0xb09a7, World 4751704) to device naa.60a98000316b6164792b457375544948 timed out: expiry time occurs 4ms in the past
2015-05-26T16:26:11.242Z cpu33:5163363)ScsiDeviceIO: 2358: Cmd(0x4124c07d0f80) 0xfe, CmdSN 0xd2e from world 4751704 to dev "naa.60a98000316b6171412b4459382d4e4a" failed H:0x5 D:0x0 P:0x0 Possible sense data: 0xa 0xd 0x2.
SOLUTION
Link to home
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
Start Free Trial
Avatar of patron

ASKER

Thanks Andrew for all you help on this...now what i need to do is either to disable VAAI or to update my HBA driver to be complaint  with min version
Avatar of patron

ASKER

Thanks