Losing pings on live migration

I currently have a 2 node OVM 3.1.1 cluster, fully patched, and have noticed that I lose network connectivity while performing a live migration. I have separate networks for VM's and for live migration. I am new to OVM so I do not know if this is typical or not. It appears that every time I migrate a VM, the client will lose connectivity to the network for anywhere from 5 - 20 seconds.

Are you talking about lost pings to the guest VM that is live migrated? If yes, that I'd suppose that to be normal behaviour, since there has to be some interruption, once the memory contents is finally synchronized between the source and the target VM server.
Depending on much RAM is used by the running VM and the speed of your network, this "outage" might vary in time, but there will surely always be some time span where pings to the VM that is being live migrated get lost.

Similar Messages

  • Live migration Vnic on hosts randomly losing connectivity HELP

    Hello Everyone,
    I am building out a new 2012 R2 cluster using VMM with converged network configuration.  I have 5 physical nics and teaming 3 of them using dynamic load balancing.  I have configured 3 virtual network adapters in host which are for management,
    cluster and Live migration.  The live migration nic loses connectivity randomly and fails migrations 50% of the time.
    Hardware is IBM blades (HS22) with Broadcom Netextreme II nics.  I have updated firmware and drivers to the latest versions.  I found a forum with something that looks very similar but this was back in November so Im guessing there is a fix. 
    http://www.hyper-v.nu/archives/mvaneijk/2013/11/vnics-and-vms-loose-connectivity-at-random-on-windows-server-2012-r2/
    Really need help with this.
    Thanks

    Hi,
    Does your cluster can pass the cluster validation test? Please install the recommended hotfixes and updates for Windows Server 2012 R2-based failover clusters update first
    then monitor again.
    More information:
    Configuring Windows Failover Cluster Networks
    http://blogs.technet.com/b/askcore/archive/2014/02/20/configuring-windows-failover-cluster-networks.aspx
    Hope this helps.
    We
    are trying to better understand customer views on social support experience, so your participation in this
    interview project would be greatly appreciated if you have time.
    Thanks for helping make community forums a great place.

  • When setting up converged network in VMM cluster and live migration virtual nics not working

    Hello Everyone,
    I am having issues setting up converged network in VMM.  I have been working with MS engineers to no avail.  I am very surprised with the expertise of the MS engineers.  They had no idea what a converged network even was.  I had way more
    experience then these guys and they said there was no escalation track so I am posting here in hopes of getting some assistance.
    Everyone including our consultants says my setup is correct. 
    What I want to do:
    I have servers with 5 nics and want to use 3 of the nics for a team and then configure cluster, live migration and host management as virtual network adapters.  I have created all my logical networks, port profile with the uplink defined as team and
    networks selected.  Created logical switch and associated portprofle.  When I deploy logical switch and create virtual network adapters the logical switch works for VMs and my management nic works as well.  Problem is that the cluster and live
    migration virtual nics do not work.  The correct Vlans get pulled in for the corresponding networks and If I run get-vmnetworkadaptervlan it shows cluster and live migration in vlans 14 and 15 which is correct.  However nics do not work at all.
    I finally decided to do this via the host in powershell and everything works fine which means this is definitely an issue with VMM.  I then imported host into VMM again but now I cannot use any of the objects I created and VMM and have to use standard
    switch.
    I am really losing faith in VMM fast. 
    Hosts are 2012 R2 and VMM is 2012 R2 all fresh builds with latest drivers
    Thanks

    Have you checked our whitepaper http://gallery.technet.microsoft.com/Hybrid-Cloud-with-NVGRE-aa6e1e9a for how to configure this through VMM?
    Are you using static IP address assignment for those vNICs?
    Are you sure your are teaming the correct physical adapters where the VLANs are trunked through the connected ports?
    Note; if you create the teaming configuration outside of VMM, and then import the hosts to VMM, then VMM will not recognize the configuration. 
    The details should be all in this whitepaper.
    -kn
    Kristian (Virtualization and some coffee: http://kristiannese.blogspot.com )

  • Failover Cluster 2008 R2 - VM lose connectivity after live migration

    Hello,
    I have a Failover Cluster with 3 server nodes running. I have 2 VMs running in one the the host without problems, but when I do a live migration of the VM to another host the VM lose network connectivity, for example if I leave a ping running, the ping command
    has 2 response, and 3 packets lost, then 1 response again, then 4 packets lost again, and so on... If I live migrate the VM to the original host, everything goes OK again.
    The same bihavior is for the 2 VMs, but I do a test with a new VM and with that new VM everything Works fine, I can live migrate it to every host.
    Any advice?
    Cristian L Ruiz

    Hi Cristian Ruiz,
    What your current host nic settings now, from you description it seems you are using the incorrect network nic design. If you are using iSCSI storage it need use the dedicate
    network in cluster.
    If your NIC teaming is iconfigured in switch independent + dynamic, please try to disable VMQ on VM Setting for narrow down the issue area.
    More information:
    VMQ Deep Dive, 1 of 3
    http://blogs.technet.com/b/networking/archive/2013/09/10/vmq-deep-dive-1-of-3.aspx
    I’m glad to be of help to you!
    We
    are trying to better understand customer views on social support experience, so your participation in this
    interview project would be greatly appreciated if you have time.
    Thanks for helping make community forums a great place.
    Hi!
    thank you for your reply!
    Yes, We are using iSCSI storage, but it has its own NICs for that (2 independant NIcs just to connect the server with the storage) and they are configured to not use those NICs to cluster communication. The team configuration is just for the LAN connectivity.
    The NIC teaming is configured using BACS4 software from a DELL server and in Smart Load Balancing and Failover (as you can see here
    http://www.micronova.com.ar/cap01.jpg). The link you passed is for Windows Server 2012 and we are running Windows Server 2008 R2, BUT as you can see in the following capture the NICs has that feature disabled
    ( http://www.micronova.com.ar/cap02.jpg ).
    One test that I'm thinking to do is to remove teaming configuration and test just with one independant NIC for LAN connection. But, I do not know if you think another choice.
    Thanks in advance.
    Cristian L Ruiz
    Sorry, another choice I'm thinking too is to update the driver versión. But the server is in production and I need to take a downtime window for test that.
    Cristian L Ruiz

  • Live Migration in OVM 3.0.2 should have an interruption or not?

    Live Migration in OVM 3.0.2 should have an interruption or not?
    I mean: I have 2 OVM Servers 3.0.2 & FibreChannel Storage
    I installed a Oracle Linux 5.6 x64 in paravirtualized Mode
    When I do a Live Migration the Virtual Machine changes in seconds to the other server with the lock image. Meanwhile I ping to the machine & Im inside the command line.
    Communication interrupt like 10 seconds or sometimes more & command line does not work for the same time
    Is that correct?
    Greetings
    Alex Dávila

    alex davila wrote:
    Right now Iam testing connectivity & when I do live migration the interruption is minimal, just 1 ping lost
    I don't know why yesterday delay secondsYou might want to talk to your networking guys to make sure that PORTFAST is enabled (if you have Cisco switches) or that you have rapid STP configured. Keep in mind that we switch the MAC address of the guest from one physical server to another. The delay you saw was your network noticing and re-routing packets to the new location.

  • Oracle vm 3.1.1 ( kernel 2.6.39-200.1.9.el5uek ) : Live migration breaks

    Hello,
    we are upgrading with last update of OVS the servers.. After we upgraded one server (called A) from 2.6.39-200.1.1.el5uek to 2.6.39-200.1.9.el5uek with certified yum repository from Oracle, the live migration not working anymore in correct mode.
    If I will migrate one guest from another server to server A or vice versa , the results are the same, the 3%-10% of packets dropped. Is it a normal behaviour , if the kernel are different ? Or are this kernel/driver/xen bugged ?
    Obviously the version of OVS are always 3.1.1 and the oracle vm the 3.1.1 build 478, and previously the live migration always worked well. No errors are visible and the job gone well.
    Kind Regards
    Edited by: user10717184 on Oct 29, 2012 12:46 AM

    I try to migrate with xm command but the problem not disappear .
    The xm command not give any result code. it finished correctly, by the way or we lost 3-10% of packets or stop pinging .
    [root@******** ~]# xm migrate -l ****UUID*** ****SERVER_OVS_NAME***
    [root@******** ~]# echo $?
    0
    Now the server have both the new kernel, but it continues to have the problem. The strange thing is that if you return to previous server OVS, the pinging restart, sometime.

  • RDS 2012 re-connection after live migration.

    Is there a way to speed up the re-connection after a live migration?
    So if i am in a vm that live migrates it feels like it hangs for about 10 seconds the reconnects and is fine..... While this is OK its not ideal. Is there a way to improve this?

    Actually 10 seconds sounds like a very long time to me. In my experience using Shared Nothing Live Migration I've seen the switch being almost instantaneous, with a continual ping possibly dropping one or two packets, and certainly quick enough that it's
    unlikely any users would notice the change. So in terms of whether it can be improved I'd say yes.
    As you can see from the technical overview here
    http://technet.microsoft.com/en-us/library/hh831435.aspx the final step is for a signal to be sent to the switch informing it of the new MAC address of the servers new destination, so I wonder if the slow switch over might be connected to that, or perhaps
    some other network issue.
    Is the network connection poor between the servers which might cause a delay during the final sync of changes between the server copies? Are you moving between subnets?

  • How to Fix: Error (10698) The virtual machine () could not be live migrated to the virtual machine host () using this cluster configuration.

    I am unable to live migrate via SCVMM 2012 R2 to one Host in our 5 node cluster.  The job fails with the errors below.
    Error (10698)
    The virtual machine () could not be live migrated to the virtual machine host () using this cluster configuration.
    Recommended Action
    Check the cluster configuration and then try the operation again.
    Information (11037)
    There currently are no network adapters with network optimization available on host.
    The host properties indicate network optimization is available as indicated in the screen shot below.
    Any guidance on things to check is appreciated.
    Thanks,
    Glenn

    Here is a snippet of the cluster log when from the current VM owner node of the failed migration:
    00000e50.000025c0::2014/02/03-13:16:07.495 INFO  [RHS] Resource Virtual Machine Configuration VMNameHere called SetResourceLockedMode. LockedModeEnabled0, LockedModeReason0.
    00000b6c.00001a9c::2014/02/03-13:16:07.495 INFO  [RCM] HandleMonitorReply: LOCKEDMODE for 'Virtual Machine Configuration VMNameHere', gen(0) result 0/0.
    00000e50.000025c0::2014/02/03-13:16:07.495 INFO  [RHS] Resource Virtual Machine VMNameHere called SetResourceLockedMode. LockedModeEnabled0, LockedModeReason0.
    00000b6c.00001a9c::2014/02/03-13:16:07.495 INFO  [RCM] HandleMonitorReply: LOCKEDMODE for 'Virtual Machine VMNameHere', gen(0) result 0/0.
    00000b6c.00001a9c::2014/02/03-13:16:07.495 INFO  [RCM] HandleMonitorReply: INMEMORY_NODELOCAL_PROPERTIES for 'Virtual Machine VMNameHere', gen(0) result 0/0.
    00000b6c.000020ec::2014/02/03-13:16:07.495 INFO  [GEM] Node 3: Sending 1 messages as a batched GEM message
    00000e50.000025c0::2014/02/03-13:16:07.495 INFO  [RES] Virtual Machine Configuration <Virtual Machine Configuration VMNameHere>: Current state 'MigrationSrcWaitForOffline', event 'MigrationSrcCompleted', result 0x8007274d
    00000e50.000025c0::2014/02/03-13:16:07.495 INFO  [RES] Virtual Machine Configuration <Virtual Machine Configuration VMNameHere>: State change 'MigrationSrcWaitForOffline' -> 'Online'
    00000e50.000025c0::2014/02/03-13:16:07.495 INFO  [RES] Virtual Machine <Virtual Machine VMNameHere>: Current state 'MigrationSrcOfflinePending', event 'MigrationSrcCompleted', result 0x8007274d
    00000e50.000025c0::2014/02/03-13:16:07.495 INFO  [RES] Virtual Machine <Virtual Machine VMNameHere>: State change 'MigrationSrcOfflinePending' -> 'Online'
    00000e50.00002080::2014/02/03-13:16:07.510 ERR   [RES] Virtual Machine <Virtual Machine VMNameHere>: Live migration of 'Virtual Machine VMNameHere' failed.
    Virtual machine migration operation for 'VMNameHere' failed at migration source 'SourceHostNameHere'. (Virtual machine ID 6901D5F8-B759-4557-8A28-E36173A14443)
    The Virtual Machine Management Service failed to establish a connection for a Virtual Machine migration with host 'DestinationHostNameHere': No connection could be made because the tar
    00000e50.00002080::2014/02/03-13:16:07.510 ERR   [RHS] Resource Virtual Machine VMNameHere has cancelled offline with error code 10061.
    00000b6c.000020ec::2014/02/03-13:16:07.510 INFO  [RCM] HandleMonitorReply: OFFLINERESOURCE for 'Virtual Machine VMNameHere', gen(0) result 0/10061.
    00000b6c.000020ec::2014/02/03-13:16:07.510 INFO  [RCM] Res Virtual Machine VMNameHere: OfflinePending -> Online( StateUnknown )
    00000b6c.000020ec::2014/02/03-13:16:07.510 INFO  [RCM] TransitionToState(Virtual Machine VMNameHere) OfflinePending-->Online.
    00000b6c.00001a9c::2014/02/03-13:16:07.510 INFO  [GEM] Node 3: Sending 1 messages as a batched GEM message
    00000b6c.000020ec::2014/02/03-13:16:07.510 INFO  [RCM] rcm::QueuedMovesHolder::VetoOffline: (VMNameHere with flags 0)
    00000b6c.000020ec::2014/02/03-13:16:07.510 INFO  [RCM] rcm::QueuedMovesHolder::RemoveGroup: (VMNameHere) GroupBeingMoved: false AllowMoveCancel: true NotifyMoveFailure: true
    00000b6c.000020ec::2014/02/03-13:16:07.510 INFO  [RCM] VMNameHere: Removed Flags 4 from StatusInformation. New StatusInformation 0
    00000b6c.000020ec::2014/02/03-13:16:07.510 INFO  [RCM] rcm::RcmGroup::CancelClusterGroupOperation: (VMNameHere)
    00000b6c.00001a9c::2014/02/03-13:16:07.510 INFO  [GEM] Node 3: Sending 1 messages as a batched GEM message
    00000b6c.000021a8::2014/02/03-13:16:07.510 INFO  [GUM] Node 3: executing request locally, gumId:3951, my action: /dm/update, # of updates: 1
    00000b6c.000021a8::2014/02/03-13:16:07.510 INFO  [GEM] Node 3: Sending 1 messages as a batched GEM message
    00000b6c.00001a9c::2014/02/03-13:16:07.510 INFO  [GEM] Node 3: Sending 1 messages as a batched GEM message
    00000b6c.000022a0::2014/02/03-13:16:07.510 INFO  [RCM] moved 0 tasks from staging set to task set.  TaskSetSize=0
    00000b6c.000022a0::2014/02/03-13:16:07.510 INFO  [RCM] rcm::RcmPriorityManager::StartGroups: [RCM] done, executed 0 tasks
    00000b6c.00000dd8::2014/02/03-13:16:07.510 INFO  [RCM] ignored non-local state Online for group VMNameHere
    00000b6c.000021a8::2014/02/03-13:16:07.526 INFO  [GUM] Node 3: executing request locally, gumId:3952, my action: /dm/update, # of updates: 1
    00000b6c.000021a8::2014/02/03-13:16:07.526 INFO  [GEM] Node 3: Sending 1 messages as a batched GEM message
    00000b6c.000018e4::2014/02/03-13:16:07.526 INFO  [RCM] HandleMonitorReply: INMEMORY_NODELOCAL_PROPERTIES for 'Virtual Machine VMNameHere', gen(0) result 0/0.
    No entry is made on the cluster log of the destination node. 
    To me this means the nodes cannot talk to each other, but I don’t know why.  
    They are on the same domain.  Their server names resolve properly and they can ping eachother both by name and IP.

  • Live Migration with Different CPU versions on the hosts, win 2012R2 Datacenter

    Hello
    This question have been asked in different forums but when I read the the thread's I feel that I get mixed answers.
    And most answers are dating from 2012 (Win 2008R2), I don't know if they are still correct in win 2012R2.
    So now I ask the question myself and hope to get at clear answer :)
    We are in the process of installing a new Hyper-V cluster using Win srv 2012 R2 Datacenter as OS.
    I'm planning to re-use some of the "old" servers from our current Hyper-V 2008 R2 cluster, removing it from the cluster and do a clean installation of 2012R2 Datacenter.
    But I will need to buy two new servers to manage this (with a new version of CPU, same brand (AMD))
    Old server: AMD Opteron(tm) Processor 6172 (12 Cores)
    New server:
    AMD Opteron™ 6344 (12-core)
    Now my question:
    Will Live Migration work between these servers in my new cluster without me doing any special settings in hyper-v or in the VM or what do I need to do to get this to work?
    /Anders

    Hi,
    It is important that all the hardware supporting Windows Server 2012 Failover Clusters be certified to work with Windows Server 2012. 
    In a cluster where all the nodes of the cluster are exactly the same, hardware migration is fairly straightforward. There are no concerns about differences in hardware, and
    especially no concerns about different capabilities of the CPUs.
    More information:
    When to Use Processor Compatibility Mode to Migrate Virtual Machines
    http://technet.microsoft.com/en-us/magazine/gg299590.aspx
    Hope this helps.
    We
    are trying to better understand customer views on social support experience, so your participation in this
    interview project would be greatly appreciated if you have time.
    Thanks for helping make community forums a great place.

  • Server 2012 r2 live migration fails with hardware error

    Hello all, we just upgraded one of our hyper v hosts from server 2012 to server 2012 r2; previously we had live replication setup between it and another box on the network which was also running server 2012. After installing server 2012 r2 when a live migration
    is attempted we get the message:
    "The virtual machine cannot be moved to the destination computer. The hardware on the destination computer is not compatible with the hardware requirements of this virtual machine. Virtual machine migration failed at migration source."
    The servers in question are both dell, currently we have a poweredge r910 running server 2012 and a poweredge r900 running server 2012 r2. The section under processor for "migrate to a physical computer using a different processor" is already checked
    and this same vm was successfully being live replicated before the upgrade to server 2012 r2. What would have changed around hardware requirements?
    We are migrating from server 2012 on the poweredge r910 to server 2012 r2 on the poweredge r900. Also When I say this was an upgrade, we did a full re install and wiped out the installation of server 2012 and installed server 2012 r2, this was not an upgrade
    installation.

    The only cause I’ve seen so far is virtual switches being named differently. I do remember that one of our VMs didn’t move, but we simply bypassed this problem, using one-time backup (VeeamZIP, more specifically).
    If it’s one-time operation you can use the same procedure for the VMs in question -> backup and restore them at new server.
    Kind regards, Leonardo.

  • Hyper-v 2012 r2 slow throughputs on network / live migrations

    Hi, maybe someone can point me in the right direction, I have 10 servers 5 Dell r210s and 5 Dell R320's, I have basically converted these servers to standalone hyper-v 2012 servers, so there is no clustering on any at the moment.
    Each server is configured with 2 1Gb nics teamed via a virtual switch, now when I copt files between server 1 and 2 for example I see 100MBs throughput, but if I copy a file to server 3 at the same time the file copy load splits the 100MBs throughput between
    the 2 copy processes. I was under the impression if I copied 2 files to 2 totally different servers the load would basically be split across the 2 nics effectively giving me 2Gbs throughput but this does not seem to be the case. I have played around with tcpip
    large send offloads, jumbo packets, disabled vmq on the cards, they are broadcoms. :-(  but it doesn't really seem to make a difference with all of these settings.
    The other issue is If I live migrate a 12Gb vm machine running only 2gb ram, effectively just an o/s it takes between 15 to 20 minutes to migrate, I have played around with the advanced settings, smb, compression, tcpip not real game changers, BUT if I shut
    town the vm and migrate it, it takes, just under 3 and a half minutes to move across.
    I am really stumped here, I am busy in a test phase of hyper-v but cant find any definitive documents relating to this stuff.

    Hi Mark,
    The servers (hyper-v 2012 r2) are all basically configured with ssvmm2012R2 where they all have teamed 1Gb pNics, into a virtual switch, then there are vNics for the Vmcloud, live migration etc.  The physical network is 2 Netgear Gs724T switches which
    are interlinked and each servers 1st nic is plugged into the switch1 and the second nic is plugged into the switch2.See Below Image)  The hyper-v port is set to independent Hyper-v load balancing. 
    The R320 servers are running raid 5 sas drives, the R210s have 1Tb drives mirrored.  The servers all are using DAS storage, we have not moved to looking at using iscsi and san is out the question at the moment.
    I am currently testing between 2x 320s and 2x R210s, I am not copying data to the vm's yets, I am basically testing the transfer between the actual hosts at the moment by copying a 4Gb file manually, After testing the live migrations I decided to test to
    see the transfer rates between the servers first, I have been playing around with the offload settings and rss, what I don't understand is yesterday, the copy between the servers was running up to 228Mbs ie (using both nics) when copying
    the file between the servers, and then a few hours later it only was copying at 50/60Mbs, but its now back at 113Mbs seemingly to be only using one nic.
    I was under the impression if you copy a file between 2 servers the nicks could use the 2gb bandwidth, but after reading many posts they say only one nic, so how did the copies get up to 2Gb yesterday. Then again if you copy files between 3 servers, then
    each copy would use one nic, basically giving you 2Gbs, but this is again not being seen.
    Regards Keith

  • Server 2012 cluster - virtual machine live migration does not work

    Hi,
    We have a hyper-v cluster with two nodes running Windows Server 2012. All the configurations are identical.
    When I try to make a Live migration from one node to the other I get an error message saying:
    Live migration of 'Virtual Machine XXXXXX' failed.
    I get no other error messages, not even in event viewer. This same happens with all of our virtual machines.
    A normal Quick migration works just fine for all of the virtual machines, so network configuration should not be an issue.
    The above error message does not provide much information.

    Hi,
    Please check whether your configuration meet live migration requirement:
    Two (or more) servers running Hyper-V that:
    Support hardware virtualization.
    Yes they support virtualization. 
    Are using processors from the same manufacturer (for example, all AMD or all Intel).
    Both Servers are identical and brand new Fujitsu-Siemens RX300S7 with the same kind of processor (Xeon E5-2620).
    Belong to either the same Active Directory domain, or to domains that trust each other.
    Both nodes are in the same domain.
    Virtual machines must be configured to use virtual hard disks or virtual Fibre Channel disks (no physical disks).
    All of the vitual machines have virtual hard disks.
    Use of a private network is recommended for live migration network traffic.
    Have tried this, but does not help.
    Requirements for live migration in a cluster:
    Windows Failover Clustering is enabled and configured.
    Yes
    Cluster Shared Volume (CSV) storage in the cluster is enabled.
    Yes
    Requirements for live migration using shared storage:
    All files that comprise a virtual machine (for example, virtual hard disks, snapshots, and configuration) are stored on an SMB share. They are all on the same CSV
    Permissions on the SMB share have been configured to grant access to the computer accounts of all servers running Hyper-V.
    Requirements for live migration with no shared infrastructure:
    No extra requirements exist.
    Also please refer to this article to check whether you have finished all preparation works for live migration:
    Virtual Machine Live Migration Overview
    http://technet.microsoft.com/en-us/library/hh831435.aspx
    Hyper-V: Using Live Migration with Cluster Shared Volumes in Windows Server 2008 R2
    http://technet.microsoft.com/en-us/library/dd446679(v=WS.10).aspx
    Configure and Use Live Migration on Non-clustered Virtual Machines
    http://technet.microsoft.com/en-us/library/jj134199.aspx
    Hope this helps!
    TechNet Subscriber Support
    If you are
    TechNet Subscription user and have any feedback on our support quality, please send your feedback
    here.
    Lawrence
    TechNet Community Support
    I have also read all of the technet articles but can't find anything that could help.

  • Live Migration Fails with error Synthetic FiberChannel Port: Failed to finish reserving resources on an VM using Windows Server 2012 R2 Hyper-V

    Hi, I'm currently experiencing a problem with some VMs in a Hyper-V 2012 R2 failover cluster using Fiber Channel adapters with Virtual SAN configured on the hyper-v hosts.
    I have read several articles about this issues like this ones:
    https://social.technet.microsoft.com/Forums/windowsserver/en-US/baca348d-fb57-4d8f-978b-f1e7282f89a1/synthetic-fibrechannel-port-failed-to-start-reserving-resources-with-error-insufficient-system?forum=winserverhyperv
    http://social.technet.microsoft.com/wiki/contents/articles/18698.hyper-v-virtual-fibre-channel-troubleshooting-guide.aspx
    But haven't been able to fix my issue.
    The Virtual SAN is configured on every hyper-v host node in the cluster. And every VM has 2 fiber channel adapters configured.
    All the World Wide Names are configured both on the FC Switch as well as the FC SAN.
    All the drivers for the FC Adapter in the Hyper-V Hosts have been updated to their latest versions.
    The strange thing is that the issue is not affecting all of the VMs, some of the VMs with FC adapters configured are live migrating just fine, others are getting this error.
    Quick migration works without problems.
    We even tried removing and creating new FC Adapters on a VM with problems, we had to configure the switch and SAN with the new WWN names and all, but ended up having the same problem.
    At first we thought is was related to the hosts, but since some VMs do work live migrating with FC adapters we tried migrating them on every host, everything worked well.
    My guess is that it has to be something related to the VMs itself but I haven't been able to figure out what is it.
    Any ideas on how to solve this is deeply appreciated.
    Thank you!
    Eduardo Rojas

    Hi Eduardo,
    How are things going ?
    Best Regards
    Elton Ji
    We
    are trying to better understand customer views on social support experience, so your participation in this
    interview project would be greatly appreciated if you have time.
    Thanks for helping make community forums a great place.

  • Hyper-V replica vs Shared Nothing Live Migration

      Shared Nothing Live Migration allows to transport your VM over the WAN without shutting it down (how much time it takes on io intensive vm is another story)
      Hyper-V replica does not allow to perform the DR switch without shutdown operation on primary site VM !
      why can't it take the VM live to the DR ?
      that's because if we use Shared Nothing across the WAN, we don't use the data that Hyper-V replica can and then it also breaks everything hyper-V replica does.
      Point is: how to take the VM to DR in running state, what is the best way to do that ?
    Shahid Roofi

    Hi Shahid,
    Hyper-V Replica is designed as a DR technology, not as a technique to move VMs. It assumes that should you require it, the source VM would probably be offline and therefore you would be powering up the passive copy from a previous point in time
    as its not a true synchronous replica copy. It does give you the added benefit to be able to run a planned failover which as you say, powers of the VM first, runs a final Sync then powers the new VM up. Obviously you cant have the duplicate copy of this VM
    running all the time at the remote site, otherwise you would have a split brain situation for network traffic.
    Like the live migration the shared nothing live migration is a technology aimed at moving a VM, but as you know offers the ability to do this without having shared storage and only requires a network connection. When initiated moves the whole
    VM, well copies the virtual drive and memory before sending machine writes to both, then only to the new VM when they both match. With regards to the speed, I assume you have the SNLM setup to compress data before sending across the wire?
    If you want a true live migration between remote sites, one way would be to have a SAN array between both sites synchronously replicating data, then stretch the Hyper-V cluster across both sites. Obviously this is a very expensive solution but perhaps
    the perfect scenario.
    Kind Regards
    Michael Coutanche
    Blog:   
    Twitter:   LinkedIn:
    Note: Posts are provided “AS IS” without warranty of any kind, either expressed or implied, including but not limited to the implied warranties of merchantability and/or fitness for a particular purpose.

  • Hyper-V guest SQL 2012 cluster live migration failure

    I have two IBM HX5 nodes connected to IBM DS5300. Hyper-V 2012 cluster was built on blades. In HV cluster was made six virtual machines, connected to DS5300 via HV Virtual SAN. These VMs was formed a guest SQL Cluster. Databases' files are placed on
    DS5300 storage and available through VM FibreChannel Adapters. IBM MPIO Module is installed on all hosts and VMs.
    SQL Server instances work without problem. But! When I try to live migrate SQL VM to another HV node an SQL Instance fails. In SQL error log I see:
    2013-06-19 10:39:44.07 spid1s      Error: 17053, Severity: 16, State: 1.
    2013-06-19 10:39:44.07 spid1s      SQLServerLogMgr::LogWriter: Operating system error 170(The requested resource is in use.) encountered.
    2013-06-19 10:39:44.07 spid1s      Write error during log flush.
    2013-06-19 10:39:44.07 spid55      Error: 9001, Severity: 21, State: 4.
    2013-06-19 10:39:44.07 spid55      The log for database 'Admin' is not available. Check the event log for related error messages. Resolve any errors and restart the database.
    2013-06-19 10:39:44.07 spid55      Database Admin was shutdown due to error 9001 in routine 'XdesRMFull::CommitInternal'. Restart for non-snapshot databases will be attempted after all connections to the database are aborted.
    2013-06-19 10:39:44.31 spid36s     Error: 17053, Severity: 16, State: 1.
    2013-06-19 10:39:44.31 spid36s     fcb::close-flush: Operating system error (null) encountered.
    2013-06-19 10:39:44.31 spid36s     Error: 17053, Severity: 16, State: 1.
    2013-06-19 10:39:44.31 spid36s     fcb::close-flush: Operating system error (null) encountered.
    2013-06-19 10:39:44.32 spid36s     Error: 17053, Severity: 16, State: 1.
    2013-06-19 10:39:44.32 spid36s     fcb::close-flush: Operating system error (null) encountered.
    2013-06-19 10:39:44.32 spid36s     Error: 17053, Severity: 16, State: 1.
    2013-06-19 10:39:44.32 spid36s     fcb::close-flush: Operating system error (null) encountered.
    2013-06-19 10:39:44.33 spid36s     Starting up database 'Admin'.
    2013-06-19 10:39:44.58 spid36s     349 transactions rolled forward in database 'Admin' (6:0). This is an informational message only. No user action is required.
    2013-06-19 10:39:44.58 spid36s     SQLServerLogMgr::FixupLogTail (failure): alignBuf 0x000000001A75D000, writeSize 0x400, filePos 0x156adc00
    2013-06-19 10:39:44.58 spid36s     blankSize 0x3c0000, blkOffset 0x1056e, fileSeqNo 1313, totBytesWritten 0x0
    2013-06-19 10:39:44.58 spid36s     fcb status 0x42, handle 0x0000000000000BC0, size 262144 pages
    2013-06-19 10:39:44.58 spid36s     Error: 17053, Severity: 16, State: 1.
    2013-06-19 10:39:44.58 spid36s     SQLServerLogMgr::FixupLogTail: Operating system error 170(The requested resource is in use.) encountered.
    2013-06-19 10:39:44.58 spid36s     Error: 5159, Severity: 24, State: 13.
    2013-06-19 10:39:44.58 spid36s     Operating system error 170(The requested resource is in use.) on file "v:\MSSQL\log\Admin\Log.ldf" during FixupLogTail.
    2013-06-19 10:39:44.58 spid36s     Error: 3414, Severity: 21, State: 1.
    2013-06-19 10:39:44.58 spid36s     An error occurred during recovery, preventing the database 'Admin' (6:0) from restarting. Diagnose the recovery errors and fix them, or restore from a known good backup. If errors are not corrected or expected,
    contact Technical Support.
    In windows system log I see a lot of warnings like this:
    - <Event xmlns="http://schemas.microsoft.com/win/2004/08/events/event">
    - <System>
      <Provider
    Name="Microsoft-Windows-Ntfs" Guid="{3FF37A1C-A68D-4D6E-8C9B-F79E8B16C482}" />
      <EventID>140</EventID>
      <Version>0</Version>
      <Level>3</Level>
      <Task>0</Task>
      <Opcode>0</Opcode>
      <Keywords>0x8000000000000008</Keywords>
      <TimeCreated
    SystemTime="2013-06-19T06:39:44.314400200Z" />
      <EventRecordID>25239</EventRecordID>
      <Correlation
    />
      <Execution
    ProcessID="4620" ThreadID="4284" />
      <Channel>System</Channel>
      <Computer>sql-node-5.local.net</Computer>
      <Security
    UserID="S-1-5-21-796845957-515967899-725345543-17066" />
      </System>
    - <EventData>
      <Data Name="VolumeId">\\?\Volume{752f0849-6201-48e9-8821-7db897a10305}</Data>
      <Data Name="DeviceName">\Device\HarddiskVolume70</Data>
      <Data Name="Error">0x80000011</Data>
      </EventData>
     </Event>
    The system failed to flush data to the transaction log. Corruption may occur in VolumeId: \\?\Volume{752f0849-6201-48e9-8821-7db897a10305}, DeviceName: \Device\HarddiskVolume70.
    ({Device Busy}
    The device is currently busy.)
    There aren't any error or warning in HV hosts.

    Hello,
    I am trying to involve someone more familiar with this topic for a further look at this issue. Sometime delay might be expected from the job transferring. Your patience is greatly appreciated.
    Thank you for your understanding and support.
    Regards,
    Fanny Liu
    If you have any feedback on our support, please click 
    here.
    Fanny Liu
    TechNet Community Support

Maybe you are looking for