NetApp snapmirror resource not getting online with VCS integration

Hi,

Explain the environment more, what are the nodes called? Have you onlined it in prod and its not working in DR, or nowhere at all?

Its complaining about encryption, have you encrypted any of passwords used in the cluster configuration?

Hi Riaan,

Thanks for the responce, This is a new deployment for Global Cluster.We have two sites Primary and DR. In the primary site we have two node vcs cluster connected netapp filer1. in the DR site we have a single node vcs cluster connected separate filer.

The replication is snapmirror over the SAN dark fibe connectivity between the sites. The replication mode is async.

Now the SnapMirror integration is pending. SnapMirror agent for VCS is installed on all nodes.

follwoing is the software informatino :

OS : Linux 6.x

SFHA : SFHA5.1SP2RP4

H/W : HP DL380Gen8

Actually during the integration we have entered the password from the vcs java console. so the password is not encrypted. I am not sure whether the password should be encrypted or not.moreover the vcsencrypt command to encrypt the password is not available.

Aslo is there any port to be opened beween vcs nodes and the filer apart from the ssh port ? coz earlier it was complaining to connect the filer. so we opened the ssh port. now it is saying about encrypt passowrd.

Any hint in this issue highly appreciated.

Thanks in advance.

Uvi

Hi,

The passwords in the configuration should be encrypted. /Should look like this in main.cf

NetAppSnapMirror sm (
                          VolumeName = unixvol
                          SnapMirrorSchedule = "0-59/5 * * *"
                          LocalFilerName = netapp1
                          LocalFilerUserName = root
                          LocalFilerPword = aoaMboDodOhmPocMe
                          RemoteFilerName = netapp2
                          RemoteFilerUserName = root
                          RemoteFilerPword = aoaMboDodOhmPocMe
                          )

The vcsencrypt should be located in the $VCS_HOME/BIN

Hi Riaan,

Thank for the support, I have encypted the passwd using vcsencrypt -agent command. However the LocalFilerUserName attribute is not availabe it seems. now my config is looing like below.

NetAppSnapMirror testSM (
                VolumeName = testvol
                SnapMirrorSchedule = "0 12 * *"
                LocalFilerName = "netapp1"
                LocalFilerPword = BLwerdPBjBPeHGpE
                RemoteFilerName = "netapp2"
                RemoteFilerPword = BLkdskBjBPeHGpE
                )

So now there is no encrypt password error. Is there any tcp port to be open from firewall as now the logs showing some socket error.

2014/10/01 10:27:58 VCS ERROR V-16-20059-1000 (node1) NetAppSnapMirror:testSM:online:ONTAPI 'system-get-version' failed on filer ddci-netapp1-ctrl2.qidc.sys. Error : in Zapi::invoke, cannot connect to socket

2014/10/01 10:27:58 VCS ERROR V-16-20059-2002 (node1) NetAppSnapMirror:testSM:online:Unable to connect to filer netapp1

Thanks in advance.

Uv

ok, well that is progress. Can you check/post what is reported in the NetAppSnapMirror logs and not just the engineA.log.

The cluster is going to try and connect to the netapp filer so it needs what ever ports you would when doing that. Can you connect manually and perform the command 'system-get-version' as it is trying?

Hi Riaan,

Thank You for the responce.

following is the message when i enters into netapp

netapp1> system-get-version

system-get-version not found.  Type '?' for a list of commands

following is the engine log message i am still getting

2014/10/02 10:17:28 VCS ERROR V-16-20059-1000 (node1) NetAppSnapMirror:testSM:online:ONTAPI 'system-get-version' failed on filer netapp1. Error : in Zapi::invoke, cannot connect to socket
2014/10/02 10:17:28 VCS ERROR V-16-20059-2002 (node1) NetAppSnapMirror:testSM:online:Unable to connect to filer netapp1

I am trying to get some help from support.

Thanks

Uv

Cool, let me know what the problem is.

Hi Ryan,

After a long time case with NetAapp the issue is got resolved. The Symantec support has provided a tool which is apitest.pl to validate the communication between nodes and filers. however the commnication is was failing so we logged a case with NetApp, The netapp support and our storage team they were trying many options in filer at last it worked. 

Now my service group is coming online properly at Primary and DR site.

I have one doupt that the data is visible at DR filer as expected in async replication, However after a failover to DR site the DR filer is in Write mode and the service group is also in online but the data written during the DR time is not visible at PR filer. 

please somebody explain whether it is normal behaviour or do i need to configure snapmirror in DR filer as well. i am not sure how the netapp snapmirror works but in hp eva ca it works.

Thanks.

Uv

Hi,

NetApp must configure it for you so that if you write in prod, its replicated (written) to DR. Ff you switch to DR, then its should replicate all the writes to PR. Once you failover to PR it should all be there.

I usually do a simple test

In PR make test_file

echo 1 >> test_file

switch to DR (using cluster or just netapp commands if you're not sure the netapp is working)

cat test_file (should show 1)

echo 2 >> test_file

switch to PR (using cluster or just netapp commands if you're not sure the netapp is working)

cat test_file (should show 1 and 2)

Hi Riaan,

Let me try it tomorrow and will update you.

Thanks.