SAN replication vs VMware SRM on?

Similar Questions

• vSphere replication vs. SRM

  What are the basic factors which would lead me to choose replication or SRM vsphere in a client environment?

  You ask a comparison of replication on Bay with SRM vs vSphere replication and SRM (SRM can use both) or ask about the differences between replication vSphere autonomous vs. SRM (with an RV or ABR)?

  If this is the first I would recommend this post: Articles VMware SRM - vs vSphere - VMware vSphere Blog - replication array-based replication

  If it's the second, VR does not include all the orchestration, test, reporting and class DR SRM operating company, but allows a protection individual VM and recovery into or through the clusters.  VR works to retrieve a few (less than 5-10) virtual machines as the recovery process must be made individually for each virtual computer. As mentioned made autonomous VR default orchestration, automation, without interruption of service analysis and account of the capabilities of MRS. VR is included in vSphere Essentials more than regulation or newer, the SRM is a VM or per CPU license (within vCloud Suite Enterprise).

• VMware SRM Reprotect fails with Peer table ID provided in the SRM of entry is incorrect

  I have the following Setup

  Site-a.

  VMware vCenter 5.1 U1

  VMware SRM 5.1.1

  NetApp SRA 2.0.1P2

  FAS3140C - Data ONTAP 8.1.2P4

  Site-B

  VMware vCenter 5.1 U1

  VMware SRM 5.1.1

  NetApp SRA 2.0.1P2

  FAS3140C - Data ONTAP 8.1.2P4

  I created a Protection Group based at the A Site containing a single VM with a single vmdk disk on an NFS volume.

  The NFS volume is synchronous to the Site-B

  I can perform a migration scheduled to Site B, and then reprotect an another planned migration back to Site A, but then when I try to restore the reprotect again so that I am ready for an another Site-B recovery fails in the first step, configure storage for reverse Direction with " " - Failed to reverse the replication for impossible on the devices. SRA command 'prepareReverseReplication' failed. Peer table provided in the entry MRS. ID is incorrect check the newspapers RS to verify the correct peer array ID. "

  I see nothing on the files or the newspaper of SRM server logs to indicate that it is.

  This occurs for each Protection Group, I create this is not isolated to just this one volume.  I also tried with VMFS iSCSI volumes and get exactly the same results.

  If I create a Protection on the Site-B group I can get back to the Site-A and will not restore to fail to Site-B.

  Initially, I look that the issue I saw is that I could do a backspace, but could not perform a second reprotect because the SnapMirrors were left in the State of evil, but now I see that the problem is that I can't make a reprotect from Site A to Site B.

  I completely uninstalled SRM in both places, removed from the SRM database on both sites and started again but always get the same question.

  I actually IBM N series-N6040 controllers and uses IBM Data ONTAP and SRA. brand got a call open with VMware and IBM but not very far.

  Has anyone seen this problem before and got a solution?

  Now, I've identified what causes this problem and can work around it until there is a fix of NetApp.

  The issue was the name of the file on the recovering Site Server had the same name as the file server on the Site protected with a prefix to the subject, which in my case the recovering Site go was named NSERIES01 and the protection of the Site was DRNSERIES01.  Don't forget that I had already performed a failover and restoration for the protection of the Site was my original recovery Site, so yes give it the site protected for this Protection Group is DRNSERIES01 and the recovering Site has NSERIES01 on this.

  Using the task Reprotect is the first step is to call the SRA with the command prepareReverseReplication, which calls reverseReplication.pl, which is trying to verify the SnapMirror is broken.  It gets the status of all the SnapMirrors run on the recovering Site, which in this case NSERIES01.  It then goes through each of them looking for a match of the local-spin-name section: volume-name of the source of the snapmirror, for example for my test group was trying to match NSERIES01:NFS_VMware_Test, at this stage of the source of the SnapMirror is DRNSERIES01:NFS_VMware_Test which is correct, but because the script uses a pattern matching test it corresponds to NSERIES01 : NFS_VMware_Test to DRNSERIES01:NFS_VMware_Test as NSERIES01:NFS_VMware_Test is contained in DRNSERIES01:NFS_VMware_Test. It then checks if the destination of the snapmirror corresponds to the peerArrayID (that is, in this case DRNSERIES01) he does not like the destination, properly, is NSERIES01, and then reports that the peerArrayID is incorrect. If there is no match on the local-spin-name: volume-name of the source of the snapmirror, then he continues to check the destination of the snapmirror, and when it finds a match, it checks if the peerArrayID is the source of the SnapMirror, and if it checks the State of the SnapMirror is broken off.

  I've never touched the issue with the first reprotect because DRNSERIES01:NFS_VMware_Test is not contained in the source of the SnapMirror (NSERIES01:NFS_VMware_Test) and so it will be the next test of verification of the DRNSERIES01:NFS_VMware_Test in the destination of the SnapMirror, that he finds and then checks the DRNSERIES01 against the destination which is also to and finally confirms that the relationship of SnapMirror is broken-off.

  I had changed the volume on DRNSERIES01 a while ago because I thought that the question could have because of the volume being names the same, but I had changed with a suffix of _Repl on the end and so the script was always corresponding to NSERIES01:NFS_VMware_Test to _ReplNSERIES01:NFS_VMware_TestDR.

  I have now set my group of test as follows:

  Protected site

  Name of the file server = NSERIES01

  Volume name = NFS_VMware_Test_HMR

  Recovery site

  File name = DRNSERIES01

  Volume name = NFS_VMware_Test_EWC

  I can now get on the Site recovery, Reprotect, restoration, Reprotect and Fail-over again and continue achieving recoveries and reprotect again and again as often as you can re - save on VHS of Scotch tape!

  If the file on the Site B server had a suffix on its name instead of a prefix, for example, it was named NSERIES01DR, or had a name completely different then I would never hit this bug in the SRA.  I'll wait for NetApp fix SRA.  In the meantime I will be renaming all my volumes on the site of recovery as well as to avoid this problem.

• "The remote server returned an error: (502) Bad Gateway" while matching Protected Site VC and recovery Site VC in VMware SRM

  Hi all

  I am trying to Setup VMware SRM 4.1 groups. I had configured hitachi usp vm storage based array replication in both sites.

  Now when I try to pair the two sites vCenter, I get an error at the end of the phase of matching as message "the remote server returned an error: (502) Bad Gateway ' on the protected site. But surprisingly, I am able to match both the vCenters on the recovering site.

  Please help me to get the problem.

  Thanks in advance,

  Sanjeev

  Run telnet server vCenter/SRM protected site to the recovering site server vCenter/SRM:

  Telnet IP_ADDRESS 80

  Telnet IP_ADDRESS 443

  You succeed?

• SAN replication for fault tolerance

  Hello world

  I hope that someone could point me in the right direction - it seems that I do not have enough review in the subject and the deadlines are too tight for me to explore different scenarios in depth...

  We have two data centers a few miles of each other connected by 100 Mbps link. Each datacenter will have 5 blades BL490 with ESX Standard accommodation about 50 VMs. ABC message a HP eva4400 SAN with SAN replication implemented. VC will be the first data center and the two data center are networked.

  SAN replication is block-level, so it seems that I can not just replicate changes, but all entries will have to be replicated. This should not be a problem because the link can support about 1.8 TB a jouret of data can be buffered.

  I am not however a vision how recovery would work in this case. We don't need instant recovery, I would say that 4 hours of recovery time is accepted if SRM fancy Automatic as DR scenario would not easily be accepted financial reasons, but your comments are welcome.

  Current idea is the following: replicate LUNS of primary site to the secondary. When disaster strikes, personal COMPUTER turns on the ESX host on the remote side and connects replicated LUNS, and then registered VMs and change my IP address.

  I understand that this seems horribly manual process and I'm pretty sure I missed some obvious pitfalls here.

  Could someone let me know which direction should I go? An article on the subject?

  This is a new installation and we prefer to put in place the basic recovery process and resize it later. I just need to have a good direction to allow this scalability.

  I thank very you much in advance!

  I think you have a good understanding of how it should work it is something I would like to implement, but use Veeam as support for replication. Here's what I would do for the IP addresses of the virtual machine. I like to keep the same IP addresses because I know that not every simple VM uses DNS for all of its applications, databases, etc. So unless you are POSITIVE you covered all the basics of your DNS, I'll try to use the same IP addresses. You just work on the part of this routing before start you your virtual machines. So make sure that when switch you or test your virtual machine failover, you make sure that if you have a virtual machine with the IP 10.10.10.10 with two Datacenters A & B, the routing would ensure that all your customers cannot talk to the datacenter B until you do this active way. In this way, you can test your failover

• Is it possible to install oracle fail-safe-kind-of (active-passive) SAN replication for oracle?

  Dear Experts,

  Is it possible to configure replication of SAN (active-passive) integrated security for oracle for oracle?

  To get the RTO < 60 seconds

  • If the primary host fails, the remote site should have the option to start the oracle database.
  • Sites are geographically remote.

  SAN storage replication, can it be used in my scenario?

  • I have no option to implement oracle Data Guard as its very expensive.

  Is that what someone achieved something similar with SAN storage replication?

  • I have oracle SE.
  • It is obvious that the all-flash backup can be quick.

  What should be the strategy to have DR site get a few seconds, if the primary site failed for some reason any

  • Daily incremental backup with RMAN. And archive backup runs for every 1 minutes
  • To have the site of DR upward and running within 60 seconds, it still requires too much time to restore the database on remote site and have archive APPLIED setpoint
  • What could be the solution with san replication? RTO is 60 seconds

  Thanks and greetings

  SAN storage replication, can it be used in my scenario?

  I have no option to implement oracle Data Guard as its very expensive.

  --> If you want to use the data that it is free of CHARGE, additional license is not required for normal data guard

  --> additional license is only for the ADG

  Is that what someone achieved something similar with SAN storage replication?

  I have oracle SE.

  It is obvious that the all-flash backup can be quick.

  --> Here, I understand that if oraclle EA is going to be expensive

  --> Again, you can use manual standby

  --> What you are looking for SAN replication, you will need to check with the storage vendor

  --> If your intention is to check with oracle storage, then this may be the correct discussion room

  What should be the strategy to have DR site get a few seconds, if the primary site failed for some reason any

  Daily incremental backup with RMAN. And archive backup runs for every 1 minutes

  To have the site of DR upward and running within 60 seconds, it still requires too much time to restore the database on remote site and have archive APPLIED setpoint

  What could be the solution with san replication? RTO is 60 seconds

  --> If your intention is to check with oracle storage, then this may be the correct discussion room

  Check in this

  https://community.Oracle.com/community/Server_%26_storage_systems/storage

• 3rd party such as san replication solution: strategy

  Dear Experts,

  Can you please share some of your ideas.

  Oracle DataGuard comes free with Oracle Enterprise Edition (EE). The only reason why a company could NOT use Oracle DataGuard is because they are not on the ERA and the cost of upgrading to EE just to use DataGuard does not justify the price.

  With 3rd party such as san replication solution. (for example hp3par)

  It is obvious that the all-flash backup can be quick.

  What should be the strategy to have DR site get a few seconds, if the primary site failed for some reason any

  1. daily incremental backup with RMAN. And archive backup runs for every 1 minutes

  2. so that the DR site facing up and running within 60 seconds, it still requires too much time to restore the database on remote site and have archive APPLIED setpoint

  What could be the solution with san replication? RTO = 60 seconds

  Thanks ivw

  Hello

  Great topic, I'll try to summarize all this in a post.

  I have worked with HP EVA for a few years and I must say that I have used many of the features of replication - HP EVA Business Copy (local replication) and HP EVA continues access (remote replication) as well - for the oracle database and other applications as well. Most storage systems use copy-on-write to create two points in time of the LUN (snapshot) image or clone LUN using cow again.

  Of course, like other people, said earlier the replication SAN is block based, in the same way with Dataguard, you have synchronous and asynchronous replication. When your data is written to the data files it usually goes in storage cache and then is pushed down the disc as soon as it will be more efficient, at the same point, replication is pushing data to the Dominican Republic. The same way with Dataguard if you use synchronous replication that you won't get you "speedy return" until the blocks of primary storage has been written for DR storage system.

  I never had problems with replication to be honest. I used a lot of copy of databases of production using local replication. For example if you want to quickly test a patch on the production system, you can create a snapshot that will only increase with the amount of changes you have made. The only concern is nowadays though we use ASM so if your diskgroup names are the same in DEV that the PROD does not work the approach that I mention.

  Regarding remote replication - because you do not have Dataguard more every time you failover to the Dominican Republic to your database will pass by instance recovery!

  Regardless of the type of replication is not the speed but the latency of the network - SAN or Dataguard, you will always have a penalty of a long replication distance. The biggest typical WAN latency factor is the speed of light, via fiber optic cable, which is about 124 miles per millisecond. Based on that you can calculate what would be the round-trip latency time.

  Indeed for many people SE installation choose DBvisit Standby to manage their databases on hold. I did the installation of the couple and it is really simple and works out of the box. However, your RPO depends on how often you switch your redologs and how many times you ship them to the backup site. You must configure dbvisit to spend with each log file (all the 5,10,15 minutes) OR configure archive_lag_target to configure OR use both at the same time. Also, you can store a copy of the archivelogs on a separate location just in case, for example, an NFS share.

  And finally my five cents on RAC on Clusters of extended - be careful with this one. Each of the nodes in the extended database cluster need access to both storage (primary and DR) systems. You must also third location to store a voting quorum disk, usually an NFS share. It is also very important to note that the latency between the nodes is of crucial importance due to global cache activity.

  Kind regards

  EVS

• Oracle Database Disaster Recovery SAN replication consistency

  Hi all

  just a quick question after a long search and not yet knowing the answer.

  What technology or configuration do I need make the eve of Oracle DB compatible on asynchronous SAN replication between cities?

  The version is Oracle Standard Edition 11 g, then the enterprise like DataGuard and GoldenGate features aren't possible?

  Entire database replication Oracle Streams don't work?

  Hello

  You can do it yourself (as everyone else used to do many years ago), to a product like dbvisit which is popular in this space (never used, but I heard good things about this). Streams can do it, but I think he uses triggers rather than redo log mining so he can have more impact on the source system.

  SAN replication is possible but probably very expensive and you should be sure it correctly handles oracle blocks - the seller must be able to tell if it is certified or not.

  See you soon,.

  Rich

• Cannot import the license key for VMware SRM

  Hi all

  We have 2 sites and have ESX4.0 3 for each site. We use VMware SRM for the DR solution, I managed with 'Recovery Plan Test', I want to import the license key for SRM, I now go with SRM Administrator's guide but I can't look for the text Licensing.LicenseKey box in the navigation pane of the window advanced settings (see attachment for details). Please give to recommend me

  Thank you & best regards

  Pham Duke of

  He's gone, the place has been changed. I had this problem myself. I do not have access to SRM right now, but I think I could just enter the key in vCenter / Home / license. Hope I remember this right.

  Best regards, Poort

• Replication of vSphere SRM - replica of the lack of virtual machines

  Hello

  I use SRM with vSphere replication and all my virtual machines are properly replicated from my site for my recovery.site as you can see:

  

  My problem is that I don't see all the replica of the VMs replicated in the inventory of the recovery site:

  

  So when I was looking for in the web client vSphere replication monitor, I saw that all the virtual machines that are in the inventory of the recovery site are all managed by SRM. And all virtual machines managed by VR finally does not appear in the inventory.

  
  

  Someone at - it never I have this problem or know the solution?

  Concerning

  Hello and welcome to the community of VMware

  The behavior you describe is planned. Between VR and SRM, is SRM work to create what is called 'space reserved VM' on the recovering site.

  If replicate you a virtual machine using vSphere replication but skip protection by SRM, there is no virtual machine being recorded on the recovering Site until you run the RV GUI recovery operation.

  Can I ask you why do you need to see the virtual machine on the recovering site? Even if it is there, you are not able to turn on.

  Best regards

  Stefan Tsonev

• With instant vmware SRM / 3rd party backup

  If you use instant vmware and third party such as veeam backup applications (and the pictures they create) it would have an impact on the operations of MRS? (Long as databases, replication is not in the same time frame that snaps being created through veeam etc.)

  Hello

  The only impact I can think is a potential growth in the replication traffic due to the increase in the amount of block changes.

  A qoute of the Administrator's guide:

  Protection and restoration of the VMS with Snapshots
  Replication on Bay supports the protection and recovery of virtual machines with the clichés, but with
  limitations.
  -You can specify a custom location to store files to capture instant delta setting workingDir in VMX
  files. SRM does not support the use of the workingDir parameter.
  -Limitations also apply if you are running versions of ESX or ESXi Server prior to version 4.1.
  If the virtual machine has several disk VMDK files, all the files on the disc must be in the same folder
  as the VMX himself file.
  If a virtual machine is connected to a Raw Disk mapping (RDM) disk device, you need to store the mapping
  file in the same folder as the VMX file. The RDM snapshots are only available if you create the map of RDM
  using Virtual Compatibility Mode.
  If you run an ESX or ESXi Server 4.1 or later, these limitations do not apply.
  vSphere replication supports the protection of VMS with snapshots, but you can retrieve only the
  last snapshot.

  Michael.

• VMware SRM case studies and see?

  Anyone know where to read stories of success for SRM implementations and see?  If this isn't the case, you have a success stories for one of the solutions of production? Please specify your environment below.

  1 SRM or the view implementation?

  2. number of hosts?

  3. storage of the seller/types?

  4 replication DR solution?

  5. What are the best practices you used successfully implemented your solution?

  6. What are your problems that you have encountered and want to share so anyone can avoid for future deployment?

  7. What do you use for networking for DR site (VLAN stretched or Re - IP)

  8. What are your recommendations for someone looking to implement the same solutions?

  Please share your solutions so that someone can determine the success rate analysis.  Thank you to enjoy holiday everyone.

  Each of my clients have a different and the best storage solutions I've worked with uses the features of VMware View composer/Linked Clone 3.0.1 and to use solutions NetApp NFS with instant volume characteristic, rapid Clone Utility 2.0 and IVMS to clone quickly and provided images of massive office so well.  But if you want to know the real world solutions, see vmware and Netapp Web sites there are a few good case studies to read.  Otherwise, someone will fill in their environment and the solutions accordingly.

  If you found this information useful, please consider awarding points to 'Correct' or 'useful '. Thank you!!!

  Kind regards

  Stefan Nguyen

  VMware vExpert 2009

  iGeek Systems Inc.

  VMware, Citrix, Microsoft Consultant

• vSphere / space recover / SAN replication

  Hello

  In my business, we have two groups (Dell PS6100 EQL) SAN with replication configured on each volume of a group of SAN. Sometimes we move on recovery of space on a specific volume of SAN.

  To do this, I need to disable replication on that volume, I execute the recovery of space on a host ESX mapped and then I reconfigure replication that will create a new set of replicas (with a full copy of first).

  During this operation (depending on the size of the volume, a few hours), if something goes wrong on the volume of SAN, we lost VMS hosted on this volume, because the first line has not yet been created.

  I could keep the replica set before disabling the replication, but the target SAN does not take into account the new volume size... I also thought about the replica cloning before neutralizing him (and destroy) but it will take a lot of time...

  Someone at - it another idea to protect data when I proceed to a recovery of the space?

  Thanks in advance

  Hello

  You are welcome.

  Re: Clone.  Yes, creation of the clone is very fast, no matter the size of the volume.

  Re: Clone replica.  Yes, it will remain if you remove the replica set.

• VMware SRM 5.8 and EMC RP

  Hi people,

  We have implemented 5.8 SRM with EMC RP 4.1 (array-based replication). Even if we validate us migration we encounter several issues as shown below.

  ERROR 1:

  2.3 stop priority 3 VMs

  Error - operation exceeded: 300 seconds

  ERROR 2:

  5. restore the hosts of sites protected from sleep mode

  Error - the object has already been disposed or was not completely created

  ERROR 3:

  6.1.2 change the location read-only

  Error - Impossible to lower source devices. SRA command 'prepareFailover' failed. Failed to logon to user of the site IP mgmt. Check the logs of the server for more details. Do not downgrade to source devices. SRA command 'prepareFailover' failed. Failed to logon to user of the site IP mgmt. Check the logs of the server for more details.

  Error "Failed opening session for the user to the site IP mgmt", seen so even the matrix to SRM Manager configuration. But the same have resolved after you have enabled the "https" to two Can MRS. Mrs. somebody please help understand the question? I tried re-record ESXi nodes to vCenter inventory, restart ESXi and other infra servers, etc. But no luck.

  Thank you!

  Best regards

  Hari Narayanan.

  Guests of protected sites are restored from sleep mode (DPM is disabled) so that storage can be removed on their part and set to read only.

  I agree with what said GSS, the problem with the LUN downgrade error is very likely with the SRA and that is why EMC. I think as the hosts of the protected site of sleep problem can be linked to that as well (not able to remove LUN?)

  About your VM stopping, you say that the virtual machine stops successfully? The default value is for MRS wait 5 minutes (300 seconds) for closure, if the virtual machine takes more time that you do not want to change the timeout value in the advanced settings. See here for more details: Site Recovery Manager 6.0 Documentation Center

• VMware SRM with ibm midsize V3700

  Hello

  I need to deploy MIS Program with ibm midsize V3700. I have to use storage replication (table). There is a feature called mirrored that allowed this replication on the remote storage. But vmware Web site I've not seen corresponding SRA. have someone deployed this kind of config? What SRA can I use?

  I thank in advance

  You must use the card replication IBM SVC storage version 2.2.1 You can download from the VMware site. For more information, see IBM Storage Host Software Solutions Information Center