Subscribe
Accepted Solution

OSSV 3.0.1, ESX 3.5 , OC5.0 - "could not read from socket"

I have the following configuration

1. OSSV 3.0.1 installed on a ESX 3.5 server

2. NetApp Host agent 2.7 installed on the same machine.

3. OC 5.0

4. I have created  a VMFS based datastore on a NEtApp lun.

5. I have a dataset with two VMs in it.

6. The dataset is not get conformed because the destination files hits the following error during conformance

"could not read from socket".

Does anyone know what could be the problem?

1. The ESX server, destination filer and OC 5.0 server are all pingable by name from each other

2. The ESX server has a VMKernel IP configured besides the service console IP.The destination filer resolves the ESX server to its service console

IP. If I make it resolve to the VMKernel IP then the baselining fails with error "cannot connect to source filer"

On ESX server

---------------------

[root@sudhaesx35 bin]# ./svinstallcheck
SnapVault home directory: '/usr/snapvault'
SnapVault database directory: '/usr/snapvault/db'
SnapVault temporary directory: '/usr/snapvault/tmp'
SnapVault trace directory: '/usr/snapvault/trace'
SnapVault Database, Temporary and Trace directories have  69% space left (3434Mb)
SnapVault service is running
SnapVault listener is running
Snapvault NDMP interface on port 10000:
   Vendor     : Netapp
   Product    : SnapVault
   Version    : 3_0_1_2011FEB17_RC
   Host       : sudhaesx35.commvault.com
   Host Id    : 66EA9DB4EBBDC608BF2665BB5A1DE3F6
   OS Type    : ESX
   OS Version : 2.4.21-47.0.1.ELvmnixEsx
   IPV6       : Not supported
   IPV4       : Ndmp Server is responding correctly
Snapvault QSM interface on port 10566:
   IPV6       : Not supported
   IPV4       : QSM Server is responding correctly
Snapvault File Server interface on port 10555:
   IPV6       : Not supported
   IPV4       : File Server is responding correctly
Validating filesystems:
   Mount point / (/dev/sda2) is suitable for backup
   Mount point /proc (none) is a special mount, unsuitable for backup
   Mount point /dev/pts (none) is a special mount, unsuitable for backup
   Mount point /proc/bus/usb (usbdevfs) is a special mount, unsuitable for backup
   Mount point /boot (/dev/sda1) is suitable for backup
   Mount point /dev/shm (none) is a special mount, unsuitable for backup
   Mount point /var/log (/dev/sda6) is suitable for backup
   Mount point /TEST1 (vega:/softdist) is a remote mount, unsuitable for backup
   Mount point /home/sudha (vega:/softdist) is a remote mount, unsuitable for backup
The vmware plugin loaded successfully.

Check Succeeded

[root@sudhaesx35 bin]# ifconfig -a
lo        Link encap:Local Loopback
          inet addr:127.0.0.1  Mask:255.0.0.0
          UP LOOPBACK RUNNING  MTU:16436  Metric:1
          RX packets:54372 errors:0 dropped:0 overruns:0 frame:0
          TX packets:54372 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:0
          RX bytes:13597013 (12.9 Mb)  TX bytes:13597013 (12.9 Mb)

vmnic0    Link encap:Ethernet  HWaddr 00:13:72:1F:56:A0
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:10805503 errors:0 dropped:0 overruns:0 frame:0
          TX packets:266986 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:964324103 (919.6 Mb)  TX bytes:148456046 (141.5 Mb)
          Interrupt:113

vswif0    Link encap:Ethernet  HWaddr 00:50:56:44:78:C5
          inet addr:172.19.100.186  Bcast:172.19.103.255  Mask:255.255.252.0
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:9375109 errors:0 dropped:0 overruns:0 frame:0
          TX packets:123865 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000
          RX bytes:774170900 (738.3 Mb)  TX bytes:69799962 (66.5 Mb)

[root@sudhaesx35 bin]# netstat -ap | grep sv

tcp        0      0 localhost.localdomain:20033 *:*                         LISTEN      5076/svcmgr

tcp        0      0 *:10566                     *:*                         LISTEN      5077/svlistener

tcp        0      0 *:10000                     *:*                         LISTEN      5077/svlistener

tcp        0      0 *:10555                     *:*                         LISTEN      5077/svlistener

tcp        0      0 localhost.localdomain:33052 localhost.localdomain:20033 ESTABLISHED 5075/svpmgr

tcp        0      0 localhost.localdomain:33053 localhost.localdomain:20033 ESTABLISHED 5077/svlistener

tcp        0      0 localhost.localdomain:20033 localhost.localdomain:33052 ESTABLISHED 5076/svcmgr

tcp        0      0 localhost.localdomain:20033 localhost.localdomain:33053 ESTABLISHED 5076/svcmgr

unix  2      [ ]         DGRAM                    94358  5077/svlistener

[root@sudhaesx35 bin]# netstat -ap | grep ntap

tcp        0      0 *:4092                      *:*                         LISTEN      5162/ntap_agent

tcp        0      0 *:4093                      *:*                         LISTEN      5162/ntap_agent

log files don't give much infromation about whats happening.

[root@sudhaesx35 bin]# ./svconfig

path.db                       /usr/snapvault/db

path.trace                    /usr/snapvault/trace

path.tmp                      /usr/snapvault/tmp

trace.enable                  TRUE

trace.lines_per_file          0

trace.files_to_keep           0

trace.level.proc_mgr          VERBOSE

trace.level.comm_mgr          VERBOSE

trace.level.svlistener        VERBOSE

trace.level.ndmp_server       VERBOSE

trace.level.qsm_server        VERBOSE

trace.level.file_server       VERBOSE

ndmp.port                     10000

ndmp.account                  root

ndmp.password

ndmp.hostid                   66EA9DB4EBBDC608BF2665BB5A1DE3F6

ndmp.hostname

qsm.accesslist.enable         FALSE

qsm.accesslist.hosts          avocado,hfiler1,hfiler2,hfiler3,hfiler4,hfiler5,hfiler6,hfiler5.testlab.commvault.com

config.bli                    HIGH

config.db_backup              BLI

vmware.vchost                 localhost

vmware.username               root

vmware.password

vmware.https.enable           TRUE

vmware.poweroff_before_ss     FALSE

vmware.backup_powered_down_vm FALSE

vmware.enable_remote_transfer 1

vmware.port.http              80

vmware.port.https             443

[root@sudhaesx35 bin]# ping hfiler5

PING hfiler5 (172.19.69.134) 56(84) bytes of data.

64 bytes from hfiler5 (172.19.69.134): icmp_seq=0 ttl=254 time=3.82 ms

64 bytes from hfiler5 (172.19.69.134): icmp_seq=1 ttl=254 time=1.84 ms

On Filer

----------

hfiler5*> options snapvault

snapvault.access             all

snapvault.enable             on

snapvault.lockvault_log_volume

snapvault.nbu.archival_snap_default on

snapvault.ossv.compression   off

snapvault.preservesnap       off

snapvault.snapshot_for_dr_backup named_snapshot_only

hfiler5*> ping -Rs sudhaesx35

64 bytes from sudhaesx35 (172.19.100.186): icmp_seq=0 ttl=63 time=18.362 ms

RR:     172.19.100.2

        sudhaesx35 (172.19.100.186)

        sudhaesx35 (172.19.100.186)

        172.19.69.4

        hfiler5 (172.19.69.134)

Any help is greatly appreciated

from snapmirror log

dst Wed Mar  7 22:02:02 EST sudhaesx35.commvault.com:app:vmware:564d0487-c8c2-aa1d-dd8 f-a80e309713d2 hfiler5:/vol/sudhaesx35/iSCSIVM-TwoDisksOnSameDS-SudhaVol6lun1 Start

dst Wed Mar  7 22:02:04 EST sudhaesx35.commvault.com:app:vmware:564d0487-c8c2-aa1d-dd8 f-a80e309713d2 hfiler5:/vol/sudhaesx35/iSCSIVM-TwoDisksOnSameDS-SudhaVol6lun1 Abort (c ould not read from socket)

hfiler5*>

Re: OSSV 3.0.1, ESX 3.5 , OC5.0 - "could not read from socket"

This could be a support related question about OnCommand but wanted to check with to see if her team is working on an answer.    If you have a NOW login you can check to see if this is covered by a Knowledgebase article?

Mike

Re: OSSV 3.0.1, ESX 3.5 , OC5.0 - "could not read from socket"

Hi there,

My team wouldn't be the right group to work on this solution.  I would recommend calling support to help you out.

Thanks,

Lisa

Re: OSSV 3.0.1, ESX 3.5 , OC5.0 - "could not read from socket"

Hi,

If this is a P1 issue, you should contact NetApp Support.  Otherwise,  If you have a NOW login, you could ask the question in the NetAppSupport Community ares for SnapX products and software where our external and internal subject matter experts may be able to help.

Mike

Re: OSSV 3.0.1, ESX 3.5 , OC5.0 - "could not read from socket"

I got this working. Turns out that the VM that I was trying to backup was not created properly. I had migrated it from a ESX 4.x server to a ESX 3.5 server. So I re-created the VM from scratch on teh ESX 3.5 server and the backups started working.

Needed to up the logging for OSSV to figure this out. It is nto documented that you need to restart the OSSV service when you change

any of the config parameters using svconfig.

Re: OSSV 3.0.1, ESX 3.5 , OC5.0 - "could not read from socket"

That's great! Was it due to Virtual Machine hardware version differences i.e.7 versus 4?

-Amir

Re: OSSV 3.0.1, ESX 3.5 , OC5.0 - "could not read from socket"

Yes that was the issue.