About aborzenkov

aborzenkov · ‎2012-03-15

Have you looked at SnapProtect? It is claimed to support application consistent VM snapshots.

aborzenkov · ‎2012-03-15

If multiple disks had failed, head should panic and aggregate should come as incomplete on the other head. In this state you can just offline and destroy it.

aborzenkov · ‎2012-03-15

It is not clear what you are trying to do. Do you have multiple disk failures in one raid group so it is no more available?

aborzenkov · ‎2012-03-15

Single group for all LUNs and all nodes in a cluster.

aborzenkov · ‎2012-03-14

Data ONTAP 8.1 dropped Filer View. You need to use System Manager if you want GUI management.

aborzenkov · ‎2012-03-14

Complete setup of the second head as standalone Make sure both heads are running the same Data ONTAP version (update if needed) and have the same licenses Install CF license on both heads Switch heads off Cable all shelves, cable interconnect Power on both heads.

aborzenkov · ‎2012-03-14

I prefer to use single initiator group for all nodes in cluster, that ensure that every LUN is always mapped to all nodes. But your way will work as wll, and is actually used by SnapDrive.

aborzenkov · ‎2012-03-14

Which version of simulator do you use? And please, use more meaningful title next time. It is useless to title your question "Netapp Ontap" when you post in "Data ONTAP" forum on NetApp site.

aborzenkov · ‎2012-03-14

Have you checked whether volume language is identical between old and new volume?

aborzenkov · ‎2012-03-14

They are not required indeed – single path for internal disks is fully supported. But you are right in that MP-HA is recommended; I can’t comment on NetApp order units structure; I would suggest you raise this issue with your NetApp sales representative. I do not think fault LED is due to missing cross connection. Usual reason for newly delivered systems is the fact, that cf is disabled. Otherwise just open support case to verify configuration.

aborzenkov · ‎2012-03-13

The former is single controller, the latter – dual controller.

aborzenkov · ‎2012-03-13

Yes, NetApp still performs user mapping and it must succeed (i.e. – valid Unix user name must be found). By default, wafl.default_unix_user is set to “pcuser” and “pcuser” exists in /etc/passwd by default. So mapping succeeds. It does not actually matter, which user CIFS user is mapped to, because this user is not used for any permissions checking. Verify that wafl.default_unix_user is set and user it is set to exists /etc/passwd and /etc/passwd is used for user name resolution.

aborzenkov · ‎2012-03-13

Hmm … this option actually applies to NFS client, accessing NTFS qtree, not to CIFS users accessing NTFS qtree. I wonder, whether setting it changes anything.

aborzenkov · ‎2012-03-13

You can’t split shelf between nodes with hardware based disk ownership. Procedure to change to software based disk ownership is described in Storage Management Guide, in chapter Changing from hardware-based to software-based disk ownership nondisruptively.

aborzenkov · ‎2012-03-13

i. DFM is separate product that is installed on Windows or Linux ii. Correct – it is Data Fabric Manager iii. I expect you will, but just try. iv. SnapManager provides application aware snapshots. For some tasks it is using and requires DFM Protection Manager. I am not aware that DFM has anything to do with reclaiming space. This is meaningless if you have NAS, as removing file automatically “reclaims” its space. And in case of LUN this can be done only from the host which knows filesystem structure and can decide which blocks are actually free. v. SnapDrive is licensed product, although you better ask your NetApp sales representative – things change too fast. For Windows 2008 you should use windows_2008 LUN type. You could also download System Manager to configure NetApp, it could make things easier initially as it is wizard based.

aborzenkov · ‎2012-03-12

Nice to hear. Then please mark question as answered.

aborzenkov · ‎2012-03-12

No, it is not possible. Clone is based on snapshot. Any change on source which happened after this snapshot had been created is invisible on clone. If you can afford full copy, you could split clone and run “snapmirror resync” and “snapmirror update”. This would completely wipe out any changes on clone since last common snapshot and resync it with source.

aborzenkov · ‎2012-03-12

For Windows 2008 it may be possible to use MS DSM for all storage systems. It is supported by NetApp, check if it supported by your other vendor. For 2003 you have to use NetApp DSM indeed. This is always tricky situation. In principle, each DSM should bind to known storage system only, so from this side it is OK. But different vendors may have different requirements regarding driver settings, even support different driver versions. You could try to find a common set of parameters that are supported. The best is to open case both by NetApp and other vendor and ask. Or you could use third party MPIO drivers like Veritas ☺

aborzenkov · ‎2012-03-12

http://support.netapp.com/NOW/cgi-bin/software/?product=Data+ONTAP&platform=FAS2020

aborzenkov · ‎2012-03-12

You could start with telling us your operating system.

aborzenkov · ‎2012-03-12

I repeat – you have to install Data ONTAP, not motherboard firmware.

aborzenkov · ‎2012-03-12

Yes, I refer to manual case. Re 2. User opens dialogue, selects timezone then gets distracted e.g. by phone call. 5 minutes later user presses Apply. Time is now set to the value when dialogue has been opened. Which is 5 minutes in the past. Even if user is not distracted, time is still reset to some value in the past.

aborzenkov · ‎2012-03-12

The latest version is 7.3.6 (strictly speaking it is 7.3.6P5 as of this writing). What you have shown is not Data ONTAP, but motherboard firmware.

aborzenkov · ‎2012-03-12

After 4a you need to “reimage” root volume by performing fake update. Perform “software install” to get full root volume content; there is no need to do subsequent “download” as long as you stay on the same version. And as long as you will be doing it anyway, any reason you stay on 7.3.2 instead of upgrading to the latest version?

aborzenkov · ‎2012-03-11

i just want to know how the filers which are in cluster mode decide to take the ownership of the disk in hardware base environment Filer connected to ESH-A owns all disks is shelf. 2ndly nowdays the hardware base disk ownership is not supported and we can convert the hardware base to software base but that required the reboot. Non-disruptive change to software based disk ownership is supported with 7.3.1 and above.

Re: Snap, snap, snap... Bleh

Re: Offline failed aggregate without panic?

Re: Offline failed aggregate without panic?

Re: using commands to create shared luns for MSCS

Re: Netapp Data Ontap

Re: How to configure Single Head to HA?

Re: using commands to create shared luns for MSCS

Re: Netapp Ontap

Re: Access denied on migrated qtree using ndmpcopy (7.2.5.1 -> 8.0.2P6)

Re: FAS2240-4 cabling question

Re: FAS3240A

Re: CIFS multi protocol

Re: CIFS multi protocol

Re: How to change the ownership of disk from one filer to another in Hard ware base disk ownership

Re: NETAPP FAS3210 and 2 x Windows 2008R2

Re: ndmpcopy missing

Re: Cloned volume reattach

Re: Multipath Software Decision

Re: ndmpcopy missing

Re: Multipath Software Decision

Re: ndmpcopy missing

Re: Feedback on NetApp System Manager 2.0R1

Re: ndmpcopy missing

Re: ndmpcopy missing

Re: How to change the ownership of disk from one filer to another in Hard ware base disk ownership