About aborzenkov

aborzenkov · ‎2013-04-10

Changing of active interface in single mode ifgrp should have no impact on host and should not cause path switch from the host point of view. Host continues to see the same target IP over the same local interface. Your configuration is invalid. Switches to which two interfaces of ifgrp are connected must have level 2 network connectivity for all VLANs on this ifgrp. I’m surprised you were able to bring ifgrp up at all; Data ONTAP should verify connectivity between ifgrp members.

aborzenkov · ‎2013-04-09

File system size will be automatically changed next time you update SnapMirror destination.

aborzenkov · ‎2013-04-08

Well ... Snapshots also "grow" when you delete files from active filesystem or overwrite them. So using host- level tools you may well see the same or even reduced total size. It is change rate that matters.

aborzenkov · ‎2013-04-08

Personally I’d consider this a bug in SMVI; it should not rely on nfs.exports.auto-update but manage /etc/exports itself. Did you report it? What was the answer?

aborzenkov · ‎2013-04-04

Hot shelf removal is not supported.

aborzenkov · ‎2013-04-03

You still need to mark interface that has to trigger failover with "ifconfig ... nfo"

aborzenkov · ‎2013-04-02

You just connect to snapshot on SnapVault system and copy file from there. There is no built-in "single file restore from SnapVault", only from local snapshots.

aborzenkov · ‎2013-03-22

If you have unix security qtree - what are permissions for top-level directory itself? You may also try to set cifs.preserve_unix_security on, this may allow CIFS clients to explicitly control Unix mode bits. See TR-3490 for details.

aborzenkov · ‎2013-03-21

umask on Unix does not set anything - it clears bits in file modes.

aborzenkov · ‎2013-03-21

Forget this. Just read You must configure the zones using Domain IDs and port numbers; you cannot use the PWWN. Case closed

aborzenkov · ‎2013-03-21

Which one is supported/recommended? Documentation does not really say which one to use, just that which end devices should be put in the same zone. Also IIRC Cisco does not even support port based zoning so I assume WWN zoning should at least be supported.

aborzenkov · ‎2013-03-20

disk assign xx.xx -s unowned

aborzenkov · ‎2013-03-20

Yes, it is possible.

aborzenkov · ‎2013-03-20

Then - Setup snapmirror between filer1 and filer2 (do not break other relationship yet); wait until it is synchronized - Break filer1 => filer3 and filer1 => filer2 - Resync filer2 => filer3

aborzenkov · ‎2013-03-20

Are you using QSM or VSM?

aborzenkov · ‎2013-03-20

You can restrict protocols for specific adapters. So you can allow CIFS only on dedicated interfaces.

aborzenkov · ‎2013-03-20

Well … the first problem is, NetApp does not officially support hot shelf unplug. Second problem is, even if you manage to move one head and half of the shelves while second head stays in takeover mode, you have no way to perform giveback now. So you will face downtime while you be moving second head from one datacenter to another. This is exactly the same downtime as for moving both heads. So you do not save any downtime. Based on information you provided so far, I do not see any sense in doing even if it were possible.

aborzenkov · ‎2013-03-20

Of course you can move head to another datacenter, but this head will probably not useful without disks. If you want to move head and disks, say so.

aborzenkov · ‎2013-03-19

Does it help for NFS issues? http://forum.commvault.com/forums/thread/19134.aspx

aborzenkov · ‎2013-03-15

This means the first SSH key offered by client is not the same as stored in filer, so it just goes on an tries next key offered. The message is debug level in OpenSSH, but someone in NetApp decided to promote it to info level. The only fix is to make sure client is using the same public key as is stored on NetApp.

aborzenkov · ‎2013-03-12

There was at least one report where booting DOT8 with disks taken directly from DOT7 resulted in prolonged outage. So if there is a possibility to use the same version on old and new controller, this is definitely recommended. It does not really take much time to update. Another consideration is that DOT update serves as good health check. At least you can be relatively sure your old configuration is OK.

aborzenkov · ‎2013-03-11

You need to move data from internal disks to external shelves because you cannot reuse FAS2040 internal disks in FAS2240. If you have mixed aggregates that contain both internal and external disks, this becomes a challenge. Once internal disks are free and not used anymore, just follow docs.

aborzenkov · ‎2013-03-11

Amount of ports does not really the main criterion. What matters, is - Both filers run compatible Data ONTAP versions, so one can access disks taken from another - Number of disks and shelves and total amount of storage do not exceed platform limit, taken in account of course how many shelves you can connect to a single loop and how many loops you can build (here is where number of ports starts to matter). If all this is checked, it is trivial enough head swap.

aborzenkov · ‎2013-03-10

There is (or at least, was) limit to how far you can increase LUN size. This is due to requirement to preserve geometry information (actually, cylinder size); so when you grow LUN, it increases number of cylinders. It is limited to 65536, so it is how far you can grow it.

aborzenkov · ‎2013-03-10

So they were used by another system indeed. You need to force assign them to local controller. disk assign xx.xx -f. Double check disk names so you won’t accidentally assign partner disks.

Re: Effect of single-mode ifgrp interface failover on iSCSI hosts?

Re: How to resize a snapmirror volume?

Re: Determining snapshot content

Re: exportfs -w /etc/exports - safe way to clean up NFS exports?

Re: Convert Metro Cluster into HA pair

Re: Question regarding LACP with VLAN creates

Re: File restore using snapvault

Re: Umask is not working on CIFS share

Re: Umask is not working on CIFS share

Re: V-Series Brocade FMC - port based or WWN based zoning?

V-Series Brocade FMC - port based or WWN based zoning?

Re: removing a disk: proper system ID is required for option -s

Re: copy volume (with active mirror relation ship) to another filer

Re: copy volume (with active mirror relation ship) to another filer

Re: copy volume (with active mirror relation ship) to another filer

Re: How can we dedicate network ports for accessing shares only?

Re: how can I physically move filers that are mirrored ?

Re: how can I physically move filers that are mirrored ?

Re: SnapProtect Tape Copy from Snapvault Secondary doesn't work

Re: Filtering or Preventing Certain Events In Messages File

Re: Head swap 31xx to 32xx

Re: Head Upgrade

Re: FAS3160 downgrade to FAS3040

Re: LUN and Volume size

Re: Adding a second DS2246 shelf to a FAS3210