Oh sure - I typically recommend to anyone who will listen that they always run their Oracle databases with an archive_log_dest that is either on a different volume, on the local machine, or ideally both. Do at least nightly manual snapshots with putting the database in hotbackup mode, etc., and then take scheduled snapshots. Depending on your transaction volume, it could actually be much faster to recover from a scheduled snapshot an hour ago with the database active, than to recover from the scheduled snapshot at midnight and run many hours of archive logs. But then I also recommend using DataGuard wherever possible in Oracle environments too. Much better to have to make a judgement call about which recovery strategy is optimal in a failure situation than to not have enough options.
Thanks,
Matt
________________________________
From: Parisi, Justin [mailto:Justin.Parisi@netapp.com] Sent: Thursday, September 04, 2008 10:21 AM To: Matthew Zito; David.Ashton@rullion.co.uk; Milazzo Giacomo Cc: Willeke, Jochen; owner-toasters@mathworks.com; toasters@mathworks.com Subject: RE: R: Noob...
Certainly.
The question is though, how important is it to you to restore quickly? :)
________________________________
From: Matthew Zito [mailto:mzito@gridapp.com] Sent: Thursday, September 04, 2008 10:17 AM To: Parisi, Justin; David.Ashton@rullion.co.uk; Milazzo Giacomo Cc: Willeke, Jochen; owner-toasters@mathworks.com; toasters@mathworks.com Subject: RE: R: Noob...
Even if it is a database, scheduled snapshots can still be useful, though it's better if the database is in hotbackup mode or whatever the non-Oracle equivalent is. A snapshot taken while the database is active is still valid, it's just that recovering to it is like the database state after a server crashes - you have to do media recovery, etc., and some transactions may have to be rolled back. Things like archive logs, etc. are perfectly valid in these scenarios as well.
Thanks,
Matt
________________________________
From: owner-toasters@mathworks.com [mailto:owner-toasters@mathworks.com] On Behalf Of Parisi, Justin Sent: Thursday, September 04, 2008 9:52 AM To: David.Ashton@rullion.co.uk; Milazzo Giacomo Cc: Willeke, Jochen; owner-toasters@mathworks.com; toasters@mathworks.com Subject: RE: R: Noob...
David,
Keep in mind that snapshots will only grow if you delete files. So if your data incorporates a lot of deletes, then you need to take that into consideration.
If this is a database, you will want to quiesce the DB to snapshot it; if the data is simple text files, you can snap it on the fly using the volume snap schedules. Giacomo mentioned setting your snap sched to 0 0 0, but if you're not using Snapmanager or other 3rd party backups and this is not a database, you'll want to allow snapshots to take place on a schedule.
________________________________
From: David.Ashton@rullion.co.uk [mailto:David.Ashton@rullion.co.uk] Sent: Thursday, September 04, 2008 9:33 AM To: Milazzo Giacomo Cc: Willeke, Jochen; owner-toasters@mathworks.com; toasters@mathworks.com Subject: Re: R: Noob...
thanks all for the excellent advice and pointers. I'll remove the snap reserve, I think the sizing is probably correct then.
The lun is mounted on a windows server that has snapdrive installed. The rate of change - varies from 3GB somedays to 80GB on others. The server runs a legacy app that uses an enormous 3.6 million flat text files (220GB), and the number of open files is in the 10,000's at any one time so it gets a battering.
Dave
Rullion Management Services Ltd is a company registered in England and Wales with company number 2797732.
"Milazzo Giacomo" G.Milazzo@sinergy.it Sent by: owner-toasters@mathworks.com
04/09/2008 14:06
To
"Willeke, Jochen" Jochen.Willeke@wincor-nixdorf.com, David.Ashton@rullion.co.uk, toasters@mathworks.com
cc
Subject
R: Noob...
A first answer came from Jochen and I've to add that if you have a volume that has to contain a LUN (that is, with NetApp, a file) the default snapshot reservation area of 20% (or less) must be removed, so the snap schedule. In your case
Filer> snap reserve vol_bond 0 Filer> snap sched vol_bond 0 0 0
This in case you use some software of Snapmanager suite (You did not tell us the usage of that lun) or the snapmirror snapshots will be placed in the active file system area having the necessary free space. Never size a volume less that 2/2.5 time the size of the contained luns.
Bye
Da: owner-toasters@mathworks.com [mailto:owner-toasters@mathworks.com] Per conto di Willeke, Jochen Inviato: giovedì 4 settembre 2008 12.54 A: David.Ashton@rullion.co.uk; toasters@mathworks.com Oggetto: RE: Noob...
Hi David,
as far as i can see, everything is as expected. When you have fractional_reserve set to 100 (%), as soon as you create snapshots on that volume, ontap will reserve 100% of the Luns size within the volume, to gurantee writes to complete successfully.
The maths show:
820GB * 0,8 (for snap reserve) = 656 GB
320GB Lun * 2 (for fractional reserve) = 640 GB
P = 640 / (656*100) = 97,56 %
Think ontap is doing some rounding, so 98% is correct.
Regards
Jochen
________________________________
From: owner-toasters@mathworks.com [mailto:owner-toasters@mathworks.com] On Behalf Of David.Ashton@rullion.co.uk Sent: Thursday, September 04, 2008 12:04 PM To: toasters@mathworks.com Subject: Noob...
Hi, please excuse my noobness, I've recently inherited a Netapp system or 3, I'm fairly new to SANs in general and netapp in particular. This isnt my full time gig (storage), its one thing amongst the 2 thousand other things I'm working on today, but I allocate a fair amount of time to figuring this stuff out.
I'm struggling with something I'm sure all netapp storage people go through, I've read the manuals, I've played around for about 6 months on the filers and I cant answer this.
I have a FAS3020. I have a volume, lets call it vol_bond. Its 820GB, 20% snap reserve. It contains one lun, which is 320GB. The space guarantee is volume, and fractional reserve is 100.
Currently, there are 5 snapshots, totalling 8GB. It is snap mirrored to another filer. Snapshots are created nightly, there are no problems with that. At weekends, after various backup activity, these grow to about 80GB, but we only keep 5 and they're deleted.
Its all up and running, everythings fine.
Except for the filer status which reports: /vol/vol_bond is full (using or reserving 98% of space and 0% of inodes, using 49% of reserve) and the df -r output is
/vol/vol_bondv8/ 687865856 670681368 17184488 335194536 /vol/vol_bondv8/ /vol/vol_bondv8/.snapshot 171966464 14140040 157826424 0 /vol/vol_bondv8/.snapshot
What am I missing? I thought I had this thing nailed but obviously not.
Dave Ashton
**********************************************************************
The information contained in this e-mail and any attachments are intended for the named recipient(s) only. It may also be privileged and confidential. If you are not an intended recipient, you must take no action as a result of receiving it, including, but not limited to copying, distributing and amending it. If this communication has been sent to you in error, please contact us immediately and do not show the communication to any other party.
Rullion shall have no liability whatsoever in respect of the content of the communication and makes no warranty as to accuracy. Any views or opinions presented are solely those of the author.
Viruses:- Although we have taken steps to ensure that this e-mail and attachments are free from viruses, we advise that in keeping with good computing practice, the recipient should ensure they are actually virus free.
**********************************************************************
________________________________________________________________________ This e-mail has been scanned for all viruses by Star. The service is powered by MessageLabs. For more information on a proactive anti-virus service working around the clock, around the globe, visit: http://www.star.net.uk ________________________________________________________________________
________________________________
WINCOR NIXDORF International GmbH Sitz der Gesellschaft: Paderborn Registergericht Paderborn HRB 3507 Geschäftsführer: Eckard Heidloff (Vorsitzender), Stefan Auerbach, Dr. Jürgen Wunram Vorsitzender des Aufsichtsrats: Karl-Heinz Stiller Steuernummer: 339/5884/0020 - Ust-ID Nr.: DE812927716 - WEEE-Reg.-Nr. DE44477193
Diese E-Mail enthält vertrauliche Informationen. Wenn Sie nicht der richtige Adressat sind oder diese E-Mail irrtümlich erhalten haben, informieren Sie bitte sofort den Absender und vernichten Sie diese E-Mail. Das unerlaubte Kopieren sowie die unbefugte Weitergabe dieser E-Mail ist nicht gestattet.
This e-mail may contain confidential information. If you are not the intended recipient (or have received this e-mail in error) please notify the sender immediately and destroy this e-mail. Any unauthorised copying, disclosure or distribution of the material in this e-mail is strictly forbidden.
________________________________________________________________________ This e-mail has been scanned for all viruses by Star. The service is powered by MessageLabs. For more information on a proactive anti-virus service working around the clock, around the globe, visit: http://www.star.net.uk ________________________________________________________________________
________________________________________________________________________ This e-mail has been scanned for all viruses by Star. The service is powered by MessageLabs. For more information on a proactive anti-virus service working around the clock, around the globe, visit: http://www.star.net.uk ________________________________________________________________________