toasters

toasters@lists.teaparty.net

13515 discussions

Instability with full filesystems?
by Brian Tao 25 Nov '97

25 Nov '97

Has anyone noticed OnTAP 4.2a crashing when the WAFL filesystem uses up all available disk space or inodes? I was doing a series of ndmpcopies to several filers, with the intent to complete fill them to capacity with as many 2GB files as I could fit. One of the four filers crashed and rebooted early this morning: Tue Nov 25 06:00:04 EST [statd]: 6:00am up 14:40 1362 NFS ops, 0 CIFS ops, 0 HTTP ops Tue Nov 25 07:00:03 EST [statd]: 7:00am up 15:40 1362 NFS ops, 0 CIFS ops, 0 HTTP ops Tue Nov 25 12:37:32 GMT [rc]: saving 59M to /etc/crash/core.0-small ("trap 13 code = 0 eip = 196aad cs = 8 eflags = 10246 ") Tue Nov 25 12:38:04 GMT [rc]: saving 258M to /etc/crash/core.0 ("trap 13 code = 0 eip = 196aad cs = 8 eflags = 10246 ") Tue Nov 25 12:38:41 GMT [wafl_lopri]: file system is full Tue Nov 25 12:38:41 GMT [rc]: (out of disk space?) Tue Nov 25 07:38:51 EST [rc]: relog syslog Tue Nov 25 07:00:03 EST [statd]: 7:00am up 15:40 1362 NFS ops, 0 CIFS ops, 0 HTTP ops Tue Nov 25 07:38:51 EST [rc]: NetApp Release 4.2a boot complete. Tue Nov 25 07:38:52 EST [asup_main]: System Notification mail sent There wasn't enough room to save the large core file, but the small one appears to be intact. I saw something similar happen on a different filer when it ran out of inodes (unfortuately, I didn't save a copy of the panic message). If the crashes are related to full filesystem conditions, I won't worry about it too much. Just curious. ;-) -- Brian Tao (BT300, taob(a)netcom.ca) "Though this be madness, yet there is method in't"

1 0

Re: F630 4.2a CIFS unable to access snapshots?
by Daniel Quinlan 25 Nov '97

25 Nov '97

Steve Losen <scl(a)sasha.acc.virginia.edu> writes: > We have a F630 running 4.2a and are having trouble accessing > snapshots with CIFS. This may be unrelated, but some NFS clients have problems using the .snapshot way of accessing snapshots. Apparently, Netapp returns the same NFS filehandle or at least the same inode number for snapshots as for the current version of the file, and this can confuse NFS clients. sodium:~ $ ls -i .profile 602274 .profile sodium:~ $ ls -i .snapshot/hourly.0/.profile 602274 .snapshot/hourly.0/.profile To work around this problem, we set up a separate NFS export of filer:/.snapshot that we mount under /snapshot/filer. That way, you can use a completely different mount point to recover a file and the inode problem is a non-issue. Since we've had people experience file corruption when using the .snapshot in their working directory, and no problems when using the separate mount point, I asked NetApp several times for an option to disable all .snapshots except for the root one. - Dan

2 1

RE: auto-mounting CIFS shares at boot time?
by Bruce Clarke 24 Nov '97

24 Nov '97

On Monday, November 24, 1997 3:07 PM, rasmus(a)bellglobal.com [SMTP:rasmus@bellglobal.com] wrote: > Please take pity on a poor Unix hack who has gotten a bunch of NT > servers > shoved onto his network. > > I am trying to get these boxes to automatically mount a CIFS share > when > they boot. So far I have only been able to get the share to come up > when > I log a user in. When the user logs out the share disappears. Not > exactly > a good thing when you have a web server service running in the > background > trying to serve up pages directly from this Netapp cifs share. My > kingdom > for /etc/vfstab for NT! > > Any tips would be appreciated. > > -Rasmus Check out a facility called "Autoexnt". It is on the MS Tech-Net CDROM and an archive for it also exists at: http://www.savilltech.com/ntfaq/sysconf.html#sysconf43 Bruce Clarke - Database Market Development Manager Network Appliance, Inc. 2770 San Tomas Expressway Santa Clara, CA 95051 USA Phone: 1-408-367-3156 FAX: 1-408-367-3151 Email: bclarke(a)netapp.com WWW: www.netapp.com

1 0

630 Quad ethernet interface?
by Wayne 23 Nov '97

23 Nov '97

Hello, We received our first NetApp, a 630 with a slew of quad ethernet interfaces in it recently. My plan was to connect a number of Sun Ultra 2's directly to the filer using crossed ethernet cables. Also, Sun sells a QFE (Quad Fast Ethernet) card that supports the "trunking" of all four ports of this card which allows an 800Mb/s (100x4xfull duplex) connection toother compatible devices (other Suns, I assume some ethernet switches, etc.) Does anyone know if the quad ethernet controllers in the 630 can be configured this way? Also, has anyone benchmarked how much data a NetApp 630 can push to half a dozen to a dozen Ultra 2's that are directly connected via 100mb/s full duplex connections? Any tuning hints? I'd definitely appreciate learning from the collective experience. -- Wayne D. Correia Critical Path Inc. tel: +1-415-543-2800 CTO 320 First Street fax: +1-415-543-2830 www.criticalpath.net San Francisco, CA 94105 gsm: +1-415-826-6000 "we handle the world's email" http://www.criticalpath.net

5 4

backing up large filers
by matthew zeier 21 Nov '97

21 Nov '97

Has anyone found a "good" solution to backing up large filers? Someone convinced me that this ndmp thing was the way to go and I bought into the whole BudTool scheme (since they're the only ones supporting ndmp). I have an issue with that too - only one vendor. However, it's been nothing but a disaster and I'm looking for something far better. We've gotten the impression that no one at PDC has bothered to test BudTool with large filers with BreeceHill stackers (PDC and NetApp recommeneded these to us). There appears to be a general lack of interopibility. And finger pointing between Network Appliance and PDC. Right now we have an issue that keeps rebooting our filers every few days. Don't even talk about restoring data... as part of our general backup policy, we do periodic restores. I believe the wording from PDC was something to the effect that BudTool is a backup management software and not geared towards restores. Suffice it to say that we rarely have a successful restore. I'm basically fishing for alternative backup solutions. We have three F540s with ~100GB disk and about 20 other boxes with ~50GB or so each, all code development and all needs to be backed up. Any ideas? We're all pretty frustrated at the lack of a solid backup solution from Network Appliance. Hell, I'd go for something proprietary so long as it worked! - mz -- matthew zeier -- mrz(a)3com.com -- 3Com EWD Engineering -- 408/764-8420 ...................................................................... "Y el mundo se mueve, mas rapido y mejor." - Fey

3 2

F630 4.2a CIFS unable to access snapshots?
by Steve Losen 20 Nov '97

20 Nov '97

We have a F630 running 4.2a and are having trouble accessing snapshots with CIFS. Our snapshot schedule is hourlies every two hours, keep 8, and keep 7 nightlies, so we usually have 15 lying around. We use both NFS and CIFS, with unix style file protection semantics. >From NT and W95, you can get into a share and use it just fine. If you browse the ~snapshot dir, you see all the hourly.* and nightly.* entries, but if you click on any of them, you get an access denied error box. If you "dir" the snapshot directory you see all the hourly.* and nightly.* entries, but if you dir any of those, they are empty. The live directory has plenty of files in it and from NFS, you can access files in all the snapshots. Again, the live filesystem works just as you would expect. It's only the snapshots that are giving us trouble. By the way, I posted an earlier problem about the set-group bit and later realized that I was not a member of my test file's group, so of course I couldn't set the set-group bit - duh. Steve Losen scl(a)virginia.edu phone: 804-924-0640 University of Virginia ITC Unix Support

2 1

Re: Remote Backups
by mds46523＠ggr.co.uk 20 Nov '97

20 Nov '97

Could you put numbers to each of the quantities you enumerate? On Nov 20, 18:29, Scott MacDonald wrote: > Subject: Remote Backups > Toasties, > > Anyone any experiences of backing up one Filer to a remote tape > (DLT) on another Filer? We're doing some testing now, and found > that remote dumps are consistently three times slower than a local > dump (restores too). Is this normal? We're also seeing about half as > many packets coming back to the source Filer, is this really all > protocol overhead? No-one else apart from me is using these Filers, > and we're nowhere near capacity in terms of CPU, Network, Disk, or > Tape I/O. > Anyone got any experiences they can share? > > Many Thanks, > > Scott. > -- > _______________________________________________________________ > > Scott MacDonald Mobile Systems International Plc > Tel: 0171 971 8798 > System Administrator Fax: 0171 971 8712 > E-Mail: scott(a)msi-uk.com > _______________________________________________________________ >-- End of excerpt from Scott MacDonald -- -Mark, TSG Unix admin and support, int 782 4412, ext +44 1438 76 4412. ... an Englishman in London ...

1 0

Remote Backups
by Scott MacDonald 20 Nov '97

20 Nov '97

Toasties, Anyone any experiences of backing up one Filer to a remote tape (DLT) on another Filer? We're doing some testing now, and found that remote dumps are consistently three times slower than a local dump (restores too). Is this normal? We're also seeing about half as many packets coming back to the source Filer, is this really all protocol overhead? No-one else apart from me is using these Filers, and we're nowhere near capacity in terms of CPU, Network, Disk, or Tape I/O. Anyone got any experiences they can share? Many Thanks, Scott. -- _______________________________________________________________ Scott MacDonald Mobile Systems International Plc Tel: 0171 971 8798 System Administrator Fax: 0171 971 8712 E-Mail: scott(a)msi-uk.com _______________________________________________________________

2 1

locking problems?
by Alexei Rodriguez 20 Nov '97

20 Nov '97

Has anybody run into locking issues with their filers? Basically, we are seeing the following problems: - Sun Solaris 2.5.1 client tries to lock a file - the lock succeeds (it is the only client/process trying to get the the lock) - the process completes and attempts to release the lock - the process hangs because the kernel never clears the lock Our filer (f540, 8MB write cache, 256MB read) is running DataONTAP 4.2a. We are using NFS v3 over udp. hard mounts, noint. So, has anybody seen something like this? The common thread that all the processes have is that they are all perl scripts. I am looking at SunSolve for clues. Seems there have been some patches that touch on the locking. We will be testing these over the next few days. Oh, and all the lock problems started when we went from 4.01c to 4.2a. Any help would be appreciated. Thanks. Alexei

1 0

Pre-purchase Contemplations...
by Marc Powell 20 Nov '97

20 Nov '97

Greetings everyone. We are very seriously considering the purchase of either 1 F520 or 2-3 F230's to serve as customer FTP/web/mail storage. We are an ISP with approximately 2000+ customers and we expect that to grow significantly in the coming year. There are a couple of things we would like to toss out there and see what people with experience with the boxes have to say. First, and this just came up, let's say you have a F520 with redundant power per shelf. Does that mean that you only have a total of 28 bays for drives? Second, what would you all say the pros and cons of using a single F520 over several F230's would be, and vice-versa, and are the boxes really as reliable as the hype? Third, is anyone aware of any plans to develop multiple file system support into the Filers? Lastly, are there any development plans to incorporate fail-over support like that in the F630 into the F520? -- Marc Powell Systems Administrator

2 1

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