Unsolved
This post is more than 5 years old
14 Posts
0
8713
February 15th, 2015 09:00
Recover point for VM plugin issue
hello,
I'm trying to create a POC for recover point for VM.
i got two sites, each site has it's own vCenter.
I've installed a recover point cluster at each site and connect them using deployment manager. ( installed the VIB on all ESXi hosts and installed 2 vRPA for each cluster)
the installer finished successfully ( after a few attempts), but i still can't see the recover point plugin in both vCenters.
i already tried to log in and out and to restart the web client and the vCenter services and it did not solved the problem.
in vSphere client i see the plugin as available but i can't access or install it.
any idea what could be the cause for that?
how can i manage my RP clusters?
thank
Gil
fadliz
153 Posts
0
February 15th, 2015 11:00
Gil, what version are you installing? The RPVM plugin is for vSphere Web client. From what I saw in the screenshot, you are using the vCenter thick client? Make sure you review the Install Guide posted on the ECN.
Thx
Sent from my iPhone
Idan
675 Posts
0
February 16th, 2015 00:00
Hi Gil,
Can you verify that the vRPA iSCSI vNICs (eth2/3, last two network card in the VM settings) are on iSCSI-1 and 2 and on vSwitch5 ?
Also, can you make sure you can ping vmk1 and vmk2 from the vRPAs ? - this can be accomplished by SSH to the vRPAs with user/password boxmgmt and navigating to [3] Diagnostics -> [5] Run internal command.
Regards,
Idan
gilte
14 Posts
0
February 16th, 2015 00:00
I'm using vSphere web client ( vCenter 5.5), i posted the vsphere client screenshot to show that my vCenter did somehow got some information about RP4VM. anyhow, after a few more reboot i now see the plugin in vCenter.
BUT i can't see my cluster, on both vCenters.
i tried to:
1. log in and out.
2. restart vSphere web client service again
3.reboot vRPAs
but it does seems to work.
when is look at my software SCSI controller i see the it is compliant but not used. can this be the issue?
i have only one host in this cluster, i'm using stranded vSwitch with no VMkerenl ports and on physical NIC.
the VMkernel ports IPs are pingable from my computer.
* i'm using a flat network with no vlans
how can i fix that? can it be the reason i can't see my cluster in the plugin? if not what could be the reason?
thank a lot!
gilte
14 Posts
0
February 16th, 2015 01:00
thanks for the quick response.
after i've posted the question i went back to my environment and now i see the cluster, not sure how.
i can see my cluster in the administration tab and I've added a datastore to the site.
but i don't think my environment is OK.
i'm trying to follow the Administration Guide PDF in order to protect a VM,
is says to choose to automatically create a new VM copy at target cluster but it's grayed out.
in the next step i get this error:
could it be related to the problems I've encountered before, or maybe its something else?
and just to make sure:
my VMkernel ports are responding to ping form the vRPAs.
the vRPA's vNIC are connected to my network using a VM network port group which is not on the same vSwitch that is using the same subnet ( it's a lab, everything is on the same network, connected to the same physical switch)
assigning a VMkernel port group to a vNIC is not possible.
Idan
675 Posts
0
February 16th, 2015 08:00
Hi Gil,
Your network config looks OK.
As for the first issue you're seeing, normally not being able to automatically create the VM at target suggests that the VM to be protected has unsupported VDMK types such as independent or shared. Can you make sure that in your case you don't have such VMDK types ?
As for the second issue, it might be an issue in splitter to RPA communication but let's focus on the first issue for now.
Regards,
Idan Kentor
idan.kentor@emc.com
Idan
675 Posts
0
February 18th, 2015 04:00
Hi Gil,
From the admin CLI (available via SSH), run the command "get_vcenter_server_credentials" and make sure the RPAs can reach that vCenter.
Thanks,
Idan
gilte
14 Posts
0
February 18th, 2015 04:00
i've connected to the cluster IP via SSH
in get system status i get this error:
ERROR: Cluster SiteB has a connectivity problem with managed array VC_
reason: No connectivity with registered storage.
details:
gilte
14 Posts
0
February 18th, 2015 04:00
I've restarted my web client service and now i can't see my cluster in the recover point cluster.
i think that something is working with the connection between the cluster and vCenter.
where can i find the connection information ? what could be the problem ?
gilte
14 Posts
0
February 18th, 2015 05:00
i ran this command and it seems to be OK.
the vCenter information is correct ( name, IP , User name etc.)
gilte
14 Posts
0
February 18th, 2015 05:00
i also ran the following command :
get_splitter_states
and i get this result :
Cluster:
SiteB:
Splitters: None
am i suppose to get something different? do i need to manually add the splitters?
Idan
675 Posts
0
February 19th, 2015 05:00
Hi Gil,
There's no need to manually add the splitters. This looks like vRPA -> vCenter communication issues.
I'll continue to work on this offline and keep the forum updated for everyone's benefit.
Regards,
Idan
Chatch_Ingersol
5 Posts
0
February 24th, 2015 08:00
I am having the same issue. Have you found any solutions? I am trying to set this up using a XtermeIO array.I can communicate with Vcenter, but not the Vcenter storage.
-chatch
fadliz
153 Posts
0
February 24th, 2015 08:00
Chatch - Is this related to the vCenter Linked mode inquiry? If so, that would explain as that mode is not currently supported.
Thx
Idan
675 Posts
0
March 24th, 2015 08:00
Yes, this one was mainly due to connectivity issues b/w the splitter and the vRPAs.
Please contact me offline and share the SR#, I would like to see if I can somehow assist with that one.
Cheers,
Idan
idan.kentor@emc.com
jhandley04291
3 Posts
0
March 24th, 2015 08:00
Has anyone been able to solve this? Emc support is still scratching their heads over the ticket i created.
thanks