About aborzenkov

aborzenkov · ‎2012-09-18

New system health command in 8.1 looks really nice (it could have better online help though ...) but I noticed it works with SAS disks only. Are there any plans to extend it to FC as well? Those are not going to disappear in near future. Also does it support other subsystems besides storage? Thank you!

aborzenkov · ‎2012-09-17

BTW this is OnCommand UM/DFM and not System Manager AFAIK. The someone better fix the message text, please

aborzenkov · ‎2012-09-17

During upgrade to 8.1.1 7-Mode I noticed interesting message after reboot into new version (who reads those messages during normal operation ... ) Mon Sep 17 11:51:22 GMT [filer:mgr.opsmgr.autoreg.norec:warning]: No SRV records found for the System Manager server, or the server is not located on this subnet. Could somebody shed light what it tried to do? I posted here because I presumed it was speaking about OnCommand server, but feel free to move this thread to another forum if I was wrong.

aborzenkov · ‎2012-09-16

If there is online aggregate located on these disks, it will result in multidisc panic (by default) and takeover. Partner will bring aggregates online because it has access to disks.

aborzenkov · ‎2012-09-16

NetApp reclaims unused block asynchronously, so if you delete large file it can indeed take some time until free space will reflect it.

aborzenkov · ‎2012-09-14

Not in case of NetApp. If it were possible to perform all configuration using CLI, it would be fine. But unfortunately you have weird mix of commands and direct file editing. And direct file editing without front end to check for errors is always bad. It is just so error prone. And you need access to files. When you are facing system having only notebook and crossover cable, you often do not even have possibility to access files. Not to mention that NetApp CLI is stone age. Really. Yes, after ten years you get to use it, but we are speaking about customers who have other tasks than learning it. Show me consistent CLI with completion support and ability to make persistent configuration changes everywhere and I will wholeheartedly agree with you.

aborzenkov · ‎2012-09-14

Thank you, I forgot that user information is not actually part of ASUP (I work almost entirely in environments where it does not matter - either SAN or NFS, where local accounts do not exist). But even root password will probably not be restored in this case.

aborzenkov · ‎2012-09-14

Yes to both. You can use any standard backup method to preserve system volume content and you can manually reconfigure system based on ASUP information. You can also use “configure” command to create configuration snapshot (you still need to backup it somewhere). There is some configuration information (notably igroups) for which location is not known. I do not know whether it will be restored by “configure” or even full vol0 backup/restore. So you may need to do some manual adjustments anyway.

aborzenkov · ‎2012-09-13

Local controller has no idea about configuration of partners interfaces. The statement “ifconfig ifX partner ifY” means that when partner in takeover mode tries to configure interface ifY, it will be running via local interface ifX. Notice that nowhere does it refer to any “partner“ statement on partner controller. Which means that failover of partner to this controller should work, but failover of this controller to partner won’t.

aborzenkov · ‎2012-09-13

Filer distributes active paths to disk across two available paths. For Fibre Channel this happens per disk; for SAS this happens per shelf (at least, that is what I have observed). There is no absolute answer to any performance question. MP-HA distributes IO across more HBAs; OTOH now you have two controllers competing for the same FC loop. This is less relevant for SAS. But I am sure that you will be disk and/or CPU bound long before you will be HBA bound.

aborzenkov · ‎2012-09-12

In this case my first concern is that interface is not up. You can test this behavior in 7.x simulator which supports HA pair simulation (pity it was dropped in 8.x).

aborzenkov · ‎2012-09-12

8.1.1 release notes mention branchcache support. It is also described in documentation.

aborzenkov · ‎2012-09-12

Actually 8.x sets root aggregate snapshot reserve to 0% by default. Could you give reference to best practice document you mention where it recommends 5% for root aggregate?

aborzenkov · ‎2012-09-11

NetApp knows nothing about file system inside LUN. It is your responsibility to calculate new LUN size which will not damage file system. NetApp will not prevent you from setting LUN size to any value.

aborzenkov · ‎2012-09-11

Yes, each controller updates own disks only. So you need to install FW on both controllers. There is really no direct way to check for background disk update progress. The only indirect way is to check for current disk firmware. Or watch /etc/messages; when no messages about disk FW update appears for 10 minutes, it is good indication that it is completed. There is no need to remove old versions. They do not take much space and DOT always uses the latest one.

aborzenkov · ‎2012-09-11

NFS does not really have notion of session. So you can see which hosts performed mount request (which does not mean that they are actually accessing anything; and it is in principle possible to access without doing mount if client knows NFS handle); and you may see current nfsd connections with netstat. But that’s it. User information is transmitted with each request, so monitoring it is basically impossible. Probably your best bet is combination of “showmount -e” from other Unix host and netstat on NetApp (look for connections to port 2049).

aborzenkov · ‎2012-09-11

You will need to reassign disks. Try “disk assign 0a.16 -f”.

aborzenkov · ‎2012-09-11

Please, paste output of command “disk show -v”

aborzenkov · ‎2012-09-11

As far as I can tell it looks fine. So what is the problem (and message is not a problem by itself)?

aborzenkov · ‎2012-09-11

Please show “storage show disk -x” output.

aborzenkov · ‎2012-09-10

Yes, it will cause filesystem corruption. You will cut off part of filesystem, possibly containing data.

aborzenkov · ‎2012-09-10

If you already has this firmware on disk, there is no need to download it again.

aborzenkov · ‎2012-09-10

If you shrink you LUN on NetApp side without adjusting filesystem size you destroy your filesystem.

aborzenkov · ‎2012-09-10

See TR-3958: sis revert_to <7.3 </vol/volname> Converts the deduplication metafiles to appropriate lower Data ONTAP version formats, currently 7.3 or 8.0. When no volume name is provided, revert_to runs on all volumes with deduplication enabled. Did you use upgrade advisor to generate procedure? Also I strongly suggest you open case with NetApp support and ask them to provide action plan for revert. Especially if you have real data there.

aborzenkov · ‎2012-09-10

Providing output of different command in wrong time is not going to help ☺ Next time if you try to add this adapter again please provide output of exactly the command I asked you.

system health support for FC HBA/shelves/disks?

Re: Automatic registration with System Manager server - what is it?

Automatic registration with System Manager server - what is it?

Re: DS 4243 single Path HA Loop Cable failure

Re: df -k command

Re: SysMan 2.0.2 - Network Configuration Checker... where?

Re: Backup configuration of FAS2020,FAS2040

Re: Backup configuration of FAS2020,FAS2040

Re: Would a partner interface function correctly in a takeover if the interface is tagged but with n...

Re: MPHA

Re: Would a partner interface function correctly in a takeover if the interface is tagged but with n...

Re: Does NetApp Support Microsoft Branchcache?

Re: aggr options nosnap=on

Re: Reduce LUN size

Re: best practice for disk firmware on ONTAP 8.0.2 P3 7-mode

Re: is there any command equivalent to cifs sessions for nfs stats/ export ?

Re: DS14MK2 AT-FCx on FAS2020

Re: DS14MK2 AT-FCx on FAS2020

Re: DS14MK2 AT-FCx on FAS2020

Re: DS14MK2 AT-FCx on FAS2020

Re: Reduce LUN size

Re: best practice for disk firmware on ONTAP 8.0.2 P3 7-mode

Re: Reduce LUN size

Re: Downgrading from 8.1 >7.3.6P6

Re: FAS2050 - QLE2464 - System not recognizing QLE2464 (4-Port) FC Card in expansion PCI-E slot.