toasters

toasters@lists.teaparty.net

3 participants
13514 discussions

RE: Ndmp vs. Nfs backups.
by Kelly Wyatt 25 Jul '00

25 Jul '00

We use DLT7000, the compression devices on the netapp. Right now the master for these backups is a sun ultra 2. He really does not do much other than schedule jobs, perform database updates, etc. He has to create the list of files to backup and has to have a lot of temporary disk space and memory to handle this task. During the backups, most (all?) of the CPU cycles come from the filers -- all 760's. Here is some info on our sun: Sun Ultra 2 UPA/SBus (UltraSPARC-II 296MHz) with 384MB of memory. We have a 70GB raid array for /usr/openv to allow space for database growth. I thought he had more memory it may have been "borrowed" for an unexpected project. We do stagger our backups so that all backups jobs are not going for resources at one time. Each filer volume is a class for a lot of reasons. Originally we tried to limit our volumes to a size that can be restored within a 4 hour window... in reality, it is probably closer to 6 hours on some volumes, and we have a couple of "legacy" volumes from pre-multi volume days that are about 150GB. We have tested restores and seem to get around 15-17gb/hour restoring data, depending on the data. Kelly -- Kelly Wyatt, Kelly.Wyatt(a)SAS.com Senior Systems Programmer Strategic Recovery SAS Institute Inc. / SAS Campus Drive / Cary, NC 27513 http://www.sas.com SAS... The Power to Know -----Original Message----- From: Jay Orr [mailto:orrjl@stl.nexen.com] Sent: Tuesday, July 25, 2000 9:16 AM To: Kelly Wyatt Cc: 'Brian Hostetter'; toasters(a)mathworks.com Subject: RE: Ndmp vs. Nfs backups. Thanks for the data, it is quite useful! What kind of storage medium are you using (i.e. DLT?, AIT?) and what is your backup hardware (i.e. what are the specs of the system you use to run the backup software on)? This is the kind of info I can't even get the sales 'drioids to get for me... ----------- Jay Orr Systems Administrator Fujitsu Nexion Inc. St. Louis, MO

1 0

RE: Ndmp vs. Nfs backups.
by Kelly Wyatt 25 Jul '00

25 Jul '00

I have not tested NFS mounted backups versus NDMP direct attach (we use these in a multiprotocol manner, and with NFS mounts you lose the CIFS acls), but I have done backups to remote NDMP storage units using NetBackup. I hope this data is useful. Following is a note I sent to our management earlier this year. This is using NetBackup. The data was collected from NetBackup and thrown through SAS to perform the calculations. Kelly ----- I have been wondering what the affect of local (direct attach drive) versus remote (over the network) backups is on the NetApps. The following reports show that difference... For both reports (Full backups first report, Incr backups second report), the LOCAL is the mean GB/hour for local backups and the REMOTE is the mean GB/hour for remote backups. Where there is no REMOTE it means that there have been no remote backups for that volume. The PERDIFF is the difference between the LOCAL and the REMOTE (REMOTE/LOCAL). The anomalies (where remote is > ) probably means that the first big backup was an incremental backup, which, in reality, is equivalent to a full. Also, there are some volumes that always backup to remote filers... and this is indeed correct. This is on 100ethernet. We are looking at setting up a private GB network for remote backups. Full Backups 1 OBS CLASS LOCAL REMOTE PERDIFF 1 17.7295 8.1000 0.45687 2 b.vol0 24.9750 13.1800 0.52773 3 b.vol1 . 13.3340 . 4 h.vol0 . 12.7200 . 5 h.vol1 21.5450 12.6000 0.58482 6 h.vol2 0.5800 0.4725 0.81466 7 s.vol0 0.7200 1.8567 2.57870 8 s.vol1 15.8200 15.3133 0.96797 9 s.vol2 0.0400 0.0550 1.37500 10 s.vol3 . 0.0467 . Incremental Backups 2 OBS CLASS LOCAL REMOTE PERDIFF 1 11.6673 7.2805 0.62401 2 b.vol0 20.6550 13.6800 0.66231 3 b.vol1 11.4783 4.2700 0.37201 4 e.vol0 15.4419 12.2767 0.79502 5 e.vol1 15.6693 11.7867 0.75221 6 h.vol0 13.5387 11.0275 0.81452 7 h.vol1 19.7640 11.5750 0.58566 8 h.vol2 0.1453 0.0300 0.20642 9 k.vol0 9.2976 7.4225 0.79832 10 k.vol1 9.6065 6.9500 0.72347 11 k.vol3 10.7789 10.5400 0.97784 12 o.vol0 15.2288 9.4967 0.62360 13 o.vol1 6.7781 4.2133 0.62161 14 o.vol2 10.7219 8.6333 0.80521 15 p.vol0 1.1365 0.0700 0.06159 16 p.vol1 13.5583 9.6750 0.71358 17 p.vol2 14.8133 12.4600 0.84113 18 p.vol3 14.1555 9.8500 0.69584 19 q.vol1 0.2250 0.3300 1.46667 20 r.vol4 11.1685 11.3500 1.01625 21 s.vol1 13.4850 8.9160 0.66118 22 s.vol2 3.3287 0.0333 0.01001 23 s.vol3 0.0336 0.0200 0.59459 -- Kelly Wyatt, Kelly.Wyatt(a)SAS.com Senior Systems Programmer Strategic Recovery SAS Institute Inc. / SAS Campus Drive / Cary, NC 27513 http://www.sas.com SAS... The Power to Know -----Original Message----- From: Brian Hostetter [mailto:BrianH@dice.com] Sent: Monday, July 24, 2000 12:47 PM To: toasters(a)mathworks.com Subject: Ndmp vs. Nfs backups. Has anyone tested the speeds of ndmp backups vs nfs? I am looking at three different backup packages, and all have fuctions that will not work with ndmp (archiving, tape duplication, cataloging). I am wondering if nfs is any where near the backup speed of ndmp. If it is close then the benefits may outweigh the speed loss. Right now I am getting about ~8M/s per drive with ndmp which is really nice.

2 1

RE: Netapp examples - replication for disaster recovery?
by Anders Ljungberg S (QDT) 25 Jul '00

25 Jul '00

Hello Jim! Yes, at Ericsson in Stockholm we use this setup for disaster/recovery reasons. But we also use the replicated data to take our backups from. I.e we dont have to load our "Frontend Filers" with backup traffic. We have 10 F760's in 5 clusters replicating there volumes to 5 other F760's on another location. 5-6TB of data. We do this twice an hour and see 10-20sec bursts of data going over. Anders Ljungberg Ericsson Business Consulting -----Original Message----- From: Jim Klun 7183 [mailto:jklun@stercomm.com] Sent: Tuesday, July 25, 2000 1:37 PM To: Jim Klun 7183 Cc: Toaster Mailing List Subject: Re: Netapp examples - replication for disaster recovery? Thanks for responses - I noted that there was little mention of replication. Does anyone out there use replication filer-to-filer for disaster recovery or fail-over? I would like filers at sister sites replicating to one another to allow faster disaster recovery. Does anyone do this? * James A. Klun Sterling Commerce 4600 Lakehurst Ct. Dublin, Oh 43016 * * jklun(a)stercomm.com 614-793-7183 voice 614-793-7092 fax * * ---- "very like a whale" ---- * * ---- Hamlet, Act 2, Scene 2 ---- * On Mon, 24 Jul 2000, Jim Klun 7183 wrote: > > > I want to explain the NAS concept as implemented via Netapps in an > upcoming meeting. > > I am looking for real world examples where snapshot saved someones lunch. > Also uses of volume replication for 7X24 and DR. > > Finally - advantages - if they exist - in running Microsoft Exchange mail > off a netatpp. > > Thanks - meetings tomorrow. > > > > > * James A. Klun Sterling Commerce 4600 Lakehurst Ct. Dublin, Oh 43016 * > * jklun(a)stercomm.com 614-793-7183 voice 614-793-7092 fax * > * ---- "very like a whale" ---- * > * ---- Hamlet, Act 2, Scene 2 ---- * > >

1 0

C720 things
by Stephen J. Wilcox 25 Jul '00

25 Jul '00

Hi, it seems the C720 software has a filelimit of 2.2M objects imposed upon it, apparently netapp are due to increase this limit in a few months but in the meantime my C720 is using less than half its drive space so I want to get the max usage! At the moment I've got 51% object hit rate and 26% byte rate.. not too bad but I suspect I can increase it. As I understand it, when the file limit is hit if it retrieves a new object it replaces one of the older objects on the disk. Now, you can set caching rules up to ignore objects with certain strings in the URL eg ? or cgi-bin.. my logic is that if you get this right you can prevent the cache from storing objects which are nearly always dynamic and therefore increase the hit ratio as the objects on the drive end up being more static .. theres a similar section for non-cacheable mime-types. Theres also a section for cookie cheats to allow you to cache static objects within cookies. My question is has anyone spent time playing with these 3 fields and does anyone have some values which they have found to improve hit ratio that I can add to my (almost empty) config? On a different thread, if you add a netcache license how do you delete it, "DISABLE" doesnt work, the only way I found was to reinstall the OS on the filer. On -another- thread, I'm running WCCP and of course all web requests show the IP on the cache not the client itself.. is there a way to spoof the client's ip? I dont have a L4 switch btw, i run WCCP from a CAT6000 with MSFC. Thanks Steve -- Stephen J. Wilcox Internet Manager, Opal Telecom http://www.opaltelecom.co.uk/ Tel: 0161 222 2000 Fax: 0161 222 2008

1 0

Netapp examples needed - soon
by Jim Klun 7183 25 Jul '00

25 Jul '00

I want to explain the NAS concept as implemented via Netapps in an upcoming meeting. I am looking for real world examples where snapshot saved someones lunch. Also uses of volume replication for 7X24 and DR. Finally - advantages - if they exist - in running Microsoft Exchange mail off a netatpp. Thanks - meetings tomorrow. * James A. Klun Sterling Commerce 4600 Lakehurst Ct. Dublin, Oh 43016 * * jklun(a)stercomm.com 614-793-7183 voice 614-793-7092 fax * * ---- "very like a whale" ---- * * ---- Hamlet, Act 2, Scene 2 ---- *

2 2

Ndmp vs. Nfs backups.
by Brian Hostetter 24 Jul '00

24 Jul '00

Has anyone tested the speeds of ndmp backups vs nfs? I am looking at three different backup packages, and all have fuctions that will not work with ndmp (archiving, tape duplication, cataloging). I am wondering if nfs is any where near the backup speed of ndmp. If it is close then the benefits may outweigh the speed loss. Right now I am getting about ~8M/s per drive with ndmp which is really nice.

1 0

Cheap upgrade
by Stephen J. Wilcox 24 Jul '00

24 Jul '00

Does anyone know if there are ways to tweak extra performance out of a C720? I'm thinking about adding more RAM, can it take more and does it improve things? Any other thoughts? Thanks Steve -- Stephen J. Wilcox Internet Manager, Opal Telecommunications Plc http://www.opaltelecom.co.uk/ Tel: 0161 222 2000

3 6

Re: Linux-NFS list recommended I post here -- RE: NetApp versus Linux
by Eyal Traitel 23 Jul '00

23 Jul '00

Bennett, let's be more explicit: It takes about 2 min. to boot, probably starting to answer NFS calls a little bit earlier (it depends on your networking config - if you use trunking, it takes a few seconds to setup the trunk). The 5 minutes is if there's writing of a core file. I don't believe that generic server with big memory that crashes will take less to write its core file also (you can choose not to write a corefile, but this is probably unwise). Anyway, let's not forget that even if it is 5 minutes with core, a generic server crash will take longer to boot for replaying journaling, which the filer needs only a few seconds for. Does anyone have numbers for generic server boot times ? eT. --- Bennett Todd <bet(a)rahul.net> wrote: > 2000-07-20-16:25:44 Bruce Sterling Woodcock: > > > +++ NetApp (F720) > > > Boots in under 2 minutes due to the use of NVRAM even after an > > > improper shutdown. [...] > > > > 2 minutes is really unlikely unless you're just using vanilla NFS > > and haven't generated a core. Think more like 5 minutes (still > > impressive). > > That is such a bummer! > > I remember only a couple of years back, the proud boast was that you > could kick the plug out of a netapp, and no more than 45 seconds > after plugging it back in, it'd be live serving NFS data once again. > I even remember a salescritter saying (I think on this list) that > they'd had customers take that boast and try and feed it to 'em, > with a line like ``Ok, let's try it, if it's not back in 45 seconds > by my watch, you're outa here'', and stood by with confidence while > the box passed the test. > > When did the time increase to 2.6-6.6 times that once-proud > 45-second figure? > > At 5 minutes, that means Netapps are no longer booting significantly > faster than generic servers. > > -Bennett > > ATTACHMENT part 2 application/pgp-signature ===== Yours, Eyal Traitel eTraitel(a)yahoo.com, Home: 972-3-5290415 (Tel Aviv) *** eTraitel - it's the new eBuzzword ! *** __________________________________________________ Do You Yahoo!? Get Yahoo! Mail � Free email you can access from anywhere! http://mail.yahoo.com/

2 1

RE: very strange hang
by Francett, Robert D (SAIC) 23 Jul '00

23 Jul '00

FYI... Console activity at time of hang enclosed. Thanks, Bob Francett/BP Exploration (Alaska) <<toaster_console>> > ---------- > From: abond@netapp.com[SMTP:abond@netapp.com] > Sent: Saturday, July 22, 2000 2:38 PM > To: JenkinJp(a)bp.com; toasters(a)mathworks.com > Subject: RE: very strange hang > > Have you got a console attached? It would be interesting to see the > console > output when this occurs. > > Regards, > > Andrew Bond > Consulting Systems Engineer > Network Appliance UK > > > -----Original Message----- > > From: Jenkinson, John P (SAIC) [mailto:JenkinJp@bp.com] > > Sent: 22 July 2000 21:36 > > To: toasters(a)mathworks.com > > Subject: very strange hang > > > > > > we've got a strange one going currently > > one of 7 netapps, is a mirror mostly holding an oracle database > > several times a day, users can't authenticate to oracle or > > can't get query > > results > > nfs serving is ok > > only way outa this is to reboot the filer > > BUT > > when a reboot command is issues the files will go almost all > > the way down > > and put > > NFS Service unavailable in the filer's LCD and will need to > > be reset or > > powered off and back on to continue > > then we run for a few hours and repeat > > looked at locks and have done the sm_mon -l trick no joy > > locks may or may not be held no difference > > we do note a few errors on the quad ethernet card which is > > connected to the > > Sun Oracle server by rolled wires > > also note a few errors on the sun on the corresponding ports > > no other items look out of the norm > > > > any similar expereinces?? > > ----------------------------------------------- > > John Jenkinson > > BP Exploration (Alaska) > > >

1 0

Re: Linux-NFS list recommended I post here -- RE: NetApp versus Linux
by Eyal Traitel 22 Jul '00

22 Jul '00

I think you need to do: rc_toggle_basic panic "this is my panic message..." in order to force a core dump at the same moment. This is for strong hearts only ! eT. --- Brian Tao <taob(a)risc.org> wrote: > On Fri, 21 Jul 2000, Bennett Todd wrote: > > > > At 5 minutes, that means Netapps are no longer booting > significantly > > faster than generic servers. > > I just tried a reboot on one of our pre-production F740's, and > the > NFS client (a Solaris machine) saw an NFS outage of 1m16s. Granted, > this doesn't take into account a savecore, and I'm not sure how to > simulate that. It would be nice to have a firmware setting to > disable > the boot up memory scan and save another 10 or 15 seconds. > -- > Brian Tao (BT300, taob(a)risc.org) > "Though this be madness, yet there is method in't" > ===== Yours, Eyal Traitel eTraitel(a)yahoo.com, Home: 972-3-5290415 (Tel Aviv) *** eTraitel - it's the new eBuzzword ! *** __________________________________________________ Do You Yahoo!? Get Yahoo! Mail � Free email you can access from anywhere! http://mail.yahoo.com/

3 2

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

1997

toasters