toasters

toasters@lists.teaparty.net

13515 discussions

ascii files convertion UNIX <--> windows ?
by Frank Bonnet 16 Jan '02

16 Jan '02

Hi After success in accessing UNIX homedir from windows ( thanks to all that respond ) I am facing a problem with ascii file conversion An ascii file created under UNIX cannot be viewded correctly under windows ( CR LF conversion ) is there an option with Ontap to convert ascii files when opening them under windows and vice-versa ? Thanks a lot. -- Frank Bonnet

4 4

RE: ascii files convertion UNIX <--> windows ?
by Mike Carroll 16 Jan '02

16 Jan '02

Here is a little alias I created for my korn shell... it uses perl.. alias rmm='perl -i -pe '\''s/\r//'\''' Then you type "rmm <filename>" and it strips off all the damn ^M's.... -----Original Message----- From: Mike Sphar [mailto:mike.sphar@peregrine.com] Sent: Wednesday, January 16, 2002 9:55 AM To: toasters(a)mathworks.com Subject: RE: ascii files convertion UNIX <--> windows ? If Windows users just need to read (not write) UNIX files sometimes, Wordpad will do that just fine. -- Mike Sphar - Sr Systems Administrator - Engineering Support Peregrine Systems, Inc. -----Original Message----- From: Steve Losen [mailto:scl@sasha.acc.virginia.edu] Sent: Wednesday, January 16, 2002 7:28 AM To: Frank Bonnet Cc: toasters(a)mathworks.com Subject: Re: ascii files convertion UNIX <--> windows ? > Hi > > After success in accessing UNIX homedir from windows > ( thanks to all that respond ) I am facing a problem > with ascii file conversion > > An ascii file created under UNIX cannot be viewded > correctly under windows ( CR LF conversion ) is there an > option with Ontap to convert ascii files when opening > them under windows and vice-versa ? > > Thanks a lot. The filer does not do any data conversion that I am aware of. In our experience we have found that folks who use both Windows and Unix do not often manipulate the same file on both systems. Usually Windows files are only manipulated from Windows and Unix files only from Unix. However, we do recommend installing NoteTab Light on Windows and using that instead of standard Windows NotePad. NoteTab accepts all three text file line terminators: LF (unix) CRLF (dos) CR (Apple) and will generate output using any of these. So you can keep a file in its original format or convert it. http://www.notetab.com On the Unix side, Solaris has the "flip" command that converts a text file from LF to CRLF and vice versa. This utility may be available on other flavors of Unix, too. If you don't have flip, here is a trivial awk command to convert unix to dos by adding CR to each line: awk '{print $0 "\r"}' unixfile > dosfile Here is a trivial sed command to convert dos to unix by stripping the CR at the end of the line. sed 's/.$//' dosfile > unixfile On Linux, vi (vim) detects the line terminator style and preserves it. And you can also force dos line terminators with the vi command :set fileformat=dos or :set ff=dos To force unix use :set fileformat=unix or :set ff=unix Steve Losen scl(a)virginia.edu phone: 434-924-0640 University of Virginia ITC Unix Support ----------------------------------------- (on rchss001) This e-mail and any files transmitted with it are the property of Chiaro Networks, Ltd., and may contain confidential and privileged material for the sole use of the intended recipient(s) to whom this e-mail is addressed. If you are not one of the named recipient(s), or otherwise have reason to believe that you have received this message in error, please notify the sender and delete all copies from your system. Any other use, retention, dissemination, forwarding, printing or copying of this e-mail is strictly prohibited. ---------------------------------------------------------

1 0

RE: ascii files convertion UNIX <--> windows ?
by Mike Sphar 16 Jan '02

16 Jan '02

If Windows users just need to read (not write) UNIX files sometimes, Wordpad will do that just fine. -- Mike Sphar - Sr Systems Administrator - Engineering Support Peregrine Systems, Inc. -----Original Message----- From: Steve Losen [mailto:scl@sasha.acc.virginia.edu] Sent: Wednesday, January 16, 2002 7:28 AM To: Frank Bonnet Cc: toasters(a)mathworks.com Subject: Re: ascii files convertion UNIX <--> windows ? > Hi > > After success in accessing UNIX homedir from windows > ( thanks to all that respond ) I am facing a problem > with ascii file conversion > > An ascii file created under UNIX cannot be viewded > correctly under windows ( CR LF conversion ) is there an > option with Ontap to convert ascii files when opening > them under windows and vice-versa ? > > Thanks a lot. The filer does not do any data conversion that I am aware of. In our experience we have found that folks who use both Windows and Unix do not often manipulate the same file on both systems. Usually Windows files are only manipulated from Windows and Unix files only from Unix. However, we do recommend installing NoteTab Light on Windows and using that instead of standard Windows NotePad. NoteTab accepts all three text file line terminators: LF (unix) CRLF (dos) CR (Apple) and will generate output using any of these. So you can keep a file in its original format or convert it. http://www.notetab.com On the Unix side, Solaris has the "flip" command that converts a text file from LF to CRLF and vice versa. This utility may be available on other flavors of Unix, too. If you don't have flip, here is a trivial awk command to convert unix to dos by adding CR to each line: awk '{print $0 "\r"}' unixfile > dosfile Here is a trivial sed command to convert dos to unix by stripping the CR at the end of the line. sed 's/.$//' dosfile > unixfile On Linux, vi (vim) detects the line terminator style and preserves it. And you can also force dos line terminators with the vi command :set fileformat=dos or :set ff=dos To force unix use :set fileformat=unix or :set ff=unix Steve Losen scl(a)virginia.edu phone: 434-924-0640 University of Virginia ITC Unix Support

1 0

Shortcuts to CIFS home directories ?
by Frank Bonnet 15 Jan '02

15 Jan '02

Hi My auth problem is solved ;-) thanks to Adam Fox <Adam.Fox(a)netapp.com> Now I have another one ... Is it possible to create a mapping file to map home directories and allow user shortcuts to point to the right home directory as it is possible under UNIX to do with NIS maps I explain I would like windows user could maps drive as follows \\toaster\loginname the mapping file (on the toaster) would looks like \user\somedir\anotherdir\anotherdir\loginname \loginame ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ ^^^^^^^^ real home directory shortcut The best way would be to use directly NIS automount map ... Another question , does OnTap will support LDAP auth ? ( NOT active directory ) if yes , when ? Thanks a lot -- Frank Bonnet

3 2

RE: NOW enhancement: filer names
by La Londe, Renee 15 Jan '02

15 Jan '02

Hi Todd, As usual thanks for the great suggestion! We have already implemented this in the 'Manage Installed Products' form http://now.netapp.com/installedproducts and will be doing so in all of the other tools over the next 2 months. Additionally, we will be adding in the system id as another search option. Thanks again for your feedback and we will keep you informed as we update the tools. Best Regards, Renee Renee La Londe Director eService Network Appliance, INC Get Answers on NOW (Netapp on the Web)! http://now.netapp.com -----Original Message----- From: Todd C. Merrill [mailto:tmerrill@mathworks.com] Sent: Monday, January 14, 2002 7:35 AM To: toasters(a)mathworks.com Subject: NOW enhancement: filer names Importance: Low Got a suggestion for an enhancement to NOW....wherever a serial number is used (cases/RMA/profiles/prococol licenses, etc.). also show the machine name. After about 5 or 6 filers, I can't keep track of this information anymore. So, I keep a little table of my own. This is kinda silly when there is a database that keeps track of things like this already... This can easily be updated by us end users, or, since autosupport mails are sent each week, that process can do it (for those filers that actually show a serial number in the output of `sysconfig` instead of a systemID). Or, autosupport can use the contents of /etc/serial if it exists to do this on our behalf. Thanks for listening. Until next time... The Mathworks, Inc. 508-647-7000 x7792 3 Apple Hill Drive, Natick, MA 01760-2098 508-647-7001 FAX tmerrill(a)mathworks.com http://www.mathworks.com ---

1 0

cifs.home_dir option and symlinks
by neil lehrer 14 Jan '02

14 Jan '02

hi, because of our home dir structure,users/a/auser, i am trying to use the home.dir option with symlinks as it has been discussed previously in toasters. ontap 6.1.1r2, nfs and cifs. i do: cifs.home_dir /vol/vol0/homelinks in homelinks i execute ln -s /vol/vol0/userstest/a/auser auser to create the symbolic link. the problem is that it works inconsistently. i get: D:\>net use * \\file1\kusr1 System error 67 has occurred. The network name cannot be found. frequently! anybody have any ideas. this is driving me nuts. thanks. -- regards +++++++++++++++++++++++++++++++++++++++++++++++++ + Neil Lehrer + + United States International Broadcasting Bureau + System Development Division + + voice 202 619-2524 + fax 202 619-3576 + nlehrer(a)ibb.gov + + " is this crisis an opportunity or just + another grab the fire extinguisher moment?" + +++++++++++++++++++++++++++++++++++++++++++++++++

1 0

NOW enhancement: filer names
by Todd C. Merrill 14 Jan '02

14 Jan '02

Got a suggestion for an enhancement to NOW....wherever a serial number is used (cases/RMA/profiles/prococol licenses, etc.). also show the machine name. After about 5 or 6 filers, I can't keep track of this information anymore. So, I keep a little table of my own. This is kinda silly when there is a database that keeps track of things like this already... This can easily be updated by us end users, or, since autosupport mails are sent each week, that process can do it (for those filers that actually show a serial number in the output of `sysconfig` instead of a systemID). Or, autosupport can use the contents of /etc/serial if it exists to do this on our behalf. Thanks for listening. Until next time... The Mathworks, Inc. 508-647-7000 x7792 3 Apple Hill Drive, Natick, MA 01760-2098 508-647-7001 FAX tmerrill(a)mathworks.com http://www.mathworks.com ---

2 1

Re: nt subst cmd
by neil lehrer 14 Jan '02

14 Jan '02

thanks. in this context is there a diff between: ln -s /vol/vol0/smalltest/k/kusr1 kusr1 and ln -s vol/vol0/smalltest/k/kusr1 kusr1 ? Steve Losen wrote: > > > i needed to use relative paths to get this to work. > > > > i have cifs.home_dir pointed to /vol/vol0/homelinks > > > > when i mounted the filer from a sun box > > > > i went into /filer/homelinks > > > > and then i did > > ln -s ../smalltest2/k/kusr1 kusr1 > > > > where smalltest2 is a qtree at the same level as homelinks and kusr1 is the uid. > > > > that's the correct way, right? > > > > obviously[after some groaning], linking to /filer//smalltest2/k/kusr1 wouldn't > > work. > > > > > > yes? > > Yes. > > This should also work on the NFS client: > > cd /filer/homelinks > ln -s /vol/vol0/smalltest/k/kusr1 kusr1 > > The important thing to remember is that the symlink must make sense > to the filer, because the filer follows it. The symlink does not > necessarily need to work properly on the NFS client. In this case, > /vol/vol0 does not exist on the NFS client, but that doesn't stop > you from creating a symlink that points to /vol/vol0. > > It's fine to use relative links like you did with ../ when the > cifs.home_dir directory is in the same volume as all the home > directories. But what if you have multiple volumes? Then you > are forced to used the /vol/vol* absolute paths for the home > directories that are in a different volume from the cifs.home_dir > directory. > > ln -s /vol/vol0/home/k/kusr1 kusr1 > > ln -s /vol/vol2/home/x/xusr1 xusr1 > > Fortunately, ln -s does not check the filename argument on the left > to be sure it exists. So you can make symlinks that point to just > about anything you want. > > Steve Losen scl(a)virginia.edu phone: 434-924-0640 > > University of Virginia ITC Unix Support -- regards +++++++++++++++++++++++++++++++++++++++++++++++++ + Neil Lehrer + + United States International Broadcasting Bureau + System Development Division + + voice 202 619-2524 + fax 202 619-3576 + nlehrer(a)ibb.gov + + " is this crisis an opportunity or just + another grab the fire extinguisher moment?" + +++++++++++++++++++++++++++++++++++++++++++++++++

1 0

Re: service process suggestion
by MCope＠isisph.com 14 Jan '02

14 Jan '02

I feel your pain. They changed RMA shipment of parts a year or more ago. The first time I had a disk failure and it didn't arrive automatically within the four hours stipulated in the support contract I went on the war path. I was told that there were excessive parts shipments that were being sent back because of false alarms so they instituted a verification process. I argued that a message stating there had been a disk "failure" can't be interpreted any other way. I would have to say this is the only complaint I have with NetApp support services.

1 0

service process suggestion
by Todd C. Merrill 14 Jan '02

14 Jan '02

Greetings. I'm finding some of the processes within service/tech support to be annoying. 1. On more than one occasion when a disk has blown, or if we've needed a part like a cluster interconnect card, tech support has contacted me, but HELD UP dispatching the part because they wanted to verify my address. Come on. It's my responsibility to keep that up to date, and if it's not, then a mis-shipped part is my fault. Further, if filers are at alternate locations, can the NOW database's capabilities be expanded to include the location (with address and/or alternate contact number) for shipment of parts. 2. To try to alleviate this shortcoming, I've put a note in my profile that encourages tech support to use the pager number I've specified in my profile to get a hold of me. It's not my primary choice for contact, but I explicitly say to use it if I don't answer my phone or email in short order or during non-business hours. This doesn't seem to work at all. Or, maybe there can be an option to "always ship to profile address" or not, if people desire a phone call verification and a delay before a part is dispatched. I don't. Thanks for listening. Until next time... The Mathworks, Inc. 508-647-7000 x7792 3 Apple Hill Drive, Natick, MA 01760-2098 508-647-7001 FAX tmerrill(a)mathworks.com http://www.mathworks.com ---

1 0

using CIFS with UNIX auth , is it possible ?
by Frank Bonnet 14 Jan '02

14 Jan '02

Hi We use a L85 as NFS server only for now for UNIX users and I plan to also use CIFS to let PC users access to their home directory from windows. Is it possible to use UNIX (NIS) auth mechanism to map a drive from windows ? If yes pointers to basic configurations would be welcome. Thanks a lot -- Frank Bonnet

1 0

XP and NetApps
by Scott Ehrlich 12 Jan '02

12 Jan '02

>Hello to all: > >We have an NIS/NFS network with NetApps for primary storage and backup, >mostly workgroup-based workstations, with a mixture of Windows boxen >ranging from 9x - XP. We also need to have clear-text password enabled >for drive mappings. > >The XP users are unable to map drives to the NetApps due to Microsoft's >changes to SMB. Thus, we have a temporary workaround of having them map >to a UNIX server instead. > >What have people with XP done to resolve this NetApps/SMB problem? How >about those people especially in workgroup-based environments? > >What pros/cons is there to a Samba-based DC solution vs a Windows-based DC >solutions? > >Our DNS is UNIX-based and we will not turn on dynamic DNS. We also >upstream our DNS info to a master primary out of our control. > >Thanks for any insight. > >Scott

4 3

Solaris IP Network Multipathing -- help needed -- peer to peer w/ F840 using Quad 100BaseT
by Surlow, Jim 11 Jan '02

11 Jan '02

Toaster Admins, Am wondering if anyone has done what I'm currently trying to do. I'm trying to setup a Sun E6500 w/a Quad Fast Ethernet Card to connect directly to an F840 w/a Quad NIC using IP Network Multipathing found in Sol 8. I'm running OnTap 6.1.2. These are connected together without a switch or hub. Each of the boxes has a separate NIC to connect to our LAN. I currently have a vif established on the filer and I have the Solaris box cooperating (no errors in the logs anymore). I can mount over the vif (and use the group IP on the "Solaris vif"). The problem is that I only seeing substantial traffic over a single ethernet interface (using "vif stat <vif name>"). And when I say "substantial", it seems that there are ping packets or such on occasion checking the up state of the connections. According to http://www.sun.com/blueprints/0801/Multipathing-updt1.pdf (page 4), it says that outbound packets should be balanced over the multipath which in my tests they clearly are not. Has anyone else come across this? My test case has 9 mount points. If this fails, I'm thinking of dividing up the mounts over the 4 ports in the QFE and mounting the exports separately, but then I get no pooling of the interfaces and no failover. Pros & Cons to this? Thanks, J. Surlow

1 0

Firewall Ports
by Noll, Kevin 11 Jan '02

11 Jan '02

Hi Does anyone know what ports need to be opened if you have a Firewall between a Linux Server and a netapp Filer. The Network is NFS Version 3 running UDP. Thanks KEVIN ******************************************************************************************** " This message contains information that may be privileged or confidential and is the property of the Cap Gemini Ernst & Young Group. It is intended only for the person to whom it is addressed. If you are not the intended recipient, you are not authorized to read, print, retain, copy, disseminate, distribute, or use this message or any part thereof. If you receive this message in error, please notify the sender immediately and delete all copies of this message ". ********************************************************************************************

2 1

SRM products
by Steven Boyer 11 Jan '02

11 Jan '02

Has anyone used Storage Resource Management products, I am looking into CA's BrightStor, IBM, Tivoli, Terllisoft, SUN HighGround SRM products. I am specifically looking for file and resource management. IE. largest homes, share, and files, duplicates, orphan, stale, and archiving (HSM) Thanks -- Steven N. Boyer UNIX/NT/NetApp Support HQUSCENTCOM General Dynamics --

1 0

Re: Netbackup error
by Steve Kappel 11 Jan '02

11 Jan '02

Hi Marcus, Have you contacted VERITAS Support? It looks like the client is not configured as an NDMP client. When you add a client to a class you tell NetBackup what kind of client it is. >From: "Bui, Marcus" <Marcus_Bui(a)aimfunds.com> >To: "'toasters(a)mathworks.com'" <toasters(a)mathworks.com> >Subject: Netbackup error >Date: Fri, 11 Jan 2002 09:33:37 -0600 > >All, > >I have Veritas Netbackup v3.4 running on Soloris. NDMP is "enable" on the >filer and the NDMP extension is installed on the Veritas master and media >servers. I was able to test the set_ndmp_attrib -verify from the Veritas >server. Netapp support verified the host and host.equiv entries is correct >on the filer. However, when the backup is started it will stay in the queue >forever. Once the job is killed this is what I got in the Veritas log. >This the last part of the log entries. > >... >... >16:53:52 [13931] <2> getsockconnected: Connect to ushoufile01 on port 712 >16:53:53 [13931] <2> getsockconnected: Connect to ushoufile01 on port 591 >16:53:54 [13931] <2> getsockconnected: Connect to ushoufile01 on port 711 >16:53:55 [13931] <2> getsockconnected: Connect to ushoufile01 on port 771 >16:53:56 [13931] <2> getsockconnected: Connect to ushoufile01 on port 728 >16:53:56 [13931] <16> getsockconnected: unable to connect() to ushoufile01, >Connection refused (146) >16:53:56 [13931] <16> bpcr_connect: Can't connect to client ushoufile01 >16:53:56 [13931] <16> log_in_errorDB: connection refused to host ushoufile01 >16:53:56 [13931] <16> log_in_errorDB: backup of client ushoufile01 exited >with status 150 (termination requested by administrator) > > >cheers, > >Marcus Bui >email: marcus_bui(a)aimfunds.com __________________________________________________________________________ Steve Kappel steve.kappel(a)veritas.com VERITAS Software (Engineering) steve(a)stevekappel.com (Personal)

1 0

Netbackup error
by Bui, Marcus 11 Jan '02

11 Jan '02

All, I have Veritas Netbackup v3.4 running on Soloris. NDMP is "enable" on the filer and the NDMP extension is installed on the Veritas master and media servers. I was able to test the set_ndmp_attrib -verify from the Veritas server. Netapp support verified the host and host.equiv entries is correct on the filer. However, when the backup is started it will stay in the queue forever. Once the job is killed this is what I got in the Veritas log. This the last part of the log entries. ... ... 16:53:52 [13931] <2> getsockconnected: Connect to ushoufile01 on port 712 16:53:53 [13931] <2> getsockconnected: Connect to ushoufile01 on port 591 16:53:54 [13931] <2> getsockconnected: Connect to ushoufile01 on port 711 16:53:55 [13931] <2> getsockconnected: Connect to ushoufile01 on port 771 16:53:56 [13931] <2> getsockconnected: Connect to ushoufile01 on port 728 16:53:56 [13931] <16> getsockconnected: unable to connect() to ushoufile01, Connection refused (146) 16:53:56 [13931] <16> bpcr_connect: Can't connect to client ushoufile01 16:53:56 [13931] <16> log_in_errorDB: connection refused to host ushoufile01 16:53:56 [13931] <16> log_in_errorDB: backup of client ushoufile01 exited with status 150 (termination requested by administrator) cheers, Marcus Bui email: marcus_bui(a)aimfunds.com

1 0

MIB variables and front panel LCD ops display
by Todd C. Merrill 10 Jan '02

10 Jan '02

What MIB variables and formula can be used to mimic the front panel LCD ops display? Summing: tnfsCalls - "The total number of NFS calls received, since since the last boot." trpcCalls - "The total number of RPC calls received, since the last boot." cifsTotalOps - "The number of CIFS operations done by the filer, since the last time the statistics were cleared." seems to be about right. But, is there a double-count here? NFS operates over RPC. Until next time... The Mathworks, Inc. 508-647-7000 x7792 3 Apple Hill Drive, Natick, MA 01760-2098 508-647-7001 FAX tmerrill(a)mathworks.com http://www.mathworks.com ---

2 1

(no subject)
by Uros Lampret 07 Jan '02

07 Jan '02

Has anybody tried using Filers on Macintosh OS 9.1 uros _____________________________ Uros Lampret Our Space Appliances d.o.o. internet: www.ourspace.si _____________________________

2 1

Solaris 8
by Uros Lampret 07 Jan '02

07 Jan '02

Hi! I was trying F840 on Solaris 8. Transfer rate with large files was great, but on smal files (2k) was not so good. I used default NFS mount options. Should I do some NFS optimization? Is there a white paper on it? Best regards, Uros _____________________________ Uros Lampret Our Space Appliances d.o.o. Tbilisijska 59 1000 Ljubljana - Slovenia tel. +386 1 2425 372 fax +386 1 2425 382 email: uros.lampret(a)ourspace.si internet: www.ourspace.si _____________________________

3 2

FC Drive failure recovery, continued...
by Dale Gass 07 Jan '02

07 Jan '02

Thanks once again to all the incredibly helpful folks... Here's my progress so far... I edited the disk label of the drive whose controller I replaced. It looks just like the others now. So I booted the system (readonly mode, thanks for that tip! :-), and the volume is actually online, and viewable, which is a major step forward. But I'm getting a lot of errors about inconsistent directory entries: Mon Jan 7 09:36:45 AST [/e3]: wafl_nfs_lookup: inconsistent directory entry {x20 0 8166461 156921645 16794490} <th.100x100.26.jpg> in {x20 0 15140401 133992804 16794490}. I have a couple of potential courses of action right now: - Copy all the data off that I can, regardless; this is obviously a good thing to do in any case. Unfortunately, the recursive copies I attempt seem to hang after a short while. Darn... - Take my drive-with-new-controller out of the set, and let the netapp attempt a rebuild onto a new drive, ignoring media errors on the second bad drive. I worry that this will create further corruption; but given the fact my current attempt has corruption, it might not be worse...? (And I might not be able to get back to where I currently am.) A cleaner (and safer) alternative might be to boot ignoring media errors, in read only mode, with my controller-replaced-drive out of the set. In read only mode, it shouldn't rebuild the set, and ignoring media errors, it might be able to access the data in degraded mode (or at least let me view what is available with that method...) - Let the netapp repair the inconsistencies; I'm not sure the best way to proceed on this one? I'm working on the tape-restore method in parallel, but anything we can get off via some creative tweaking, would be worth the try... I think I'll try the ignore-media-errors and read-only-mode thing next, to see what that view of the world is like. -dale

1 1

RE:
by Mitko Blazeski 07 Jan '02

07 Jan '02

Hi Uros I was running a F740 with 600 Mac clients on the newspaper I was working at. Sadly the filer dosen´t support Appletalk so we bought an program called Ethershare from Helios. It worked fine and the performance was stunning. The bad thing is that the software is rather expensive. Mitko Blazeski Professional Services CS Engineer ProAct Datasystem AB Djupdalsvägen 7, SE-191 24 Stockholm -----Original Message----- From: Uros Lampret [mailto:uros.lampret@ourspace.si] Sent: den 7 januari 2002 08:08 To: toasters(a)mathworks.com Subject: Has anybody tried using Filers on Macintosh OS 9.1 uros _____________________________ Uros Lampret Our Space Appliances d.o.o. internet: www.ourspace.si <http://www.ourspace.si> _____________________________

1 0

Bogus tape write rate from sysstat
by Chris Thompson 04 Jan '02

04 Jan '02

Has anyone else noticed this? Doing "systat -u 10" during a dump I see occasional obviously bogus values in the Tape Write column, e.g. CPU Total Net kB/s Disk kB/s Tape kB/s Cache Cache CP CP Disk ops/s in out read write read write age hit time ty util 15% 12 7 5 6844 66 0 6438 >60 99% 3% T 44% 18% 3 1 2 8172 0 0 7651 >60 99% 0% - 51% 21% 43 597 617 7822 83 0 6645 >60 99% 3% T 51% 14% 12 3 8 5112 917 0 4511 >60 98% 13% T 32% 19% 12 8 3 8530 78 0 1844674406949348 >60 100% 6% T 61% 20% 7 4 4 8344 83 0 7490 >60 100% 12% T 64% 21% 6 2 10 8671 81 0 7858 >60 100% 7% T 65% 18% 8 4 14 7621 86 0 6896 >60 99% 5% T 57% I think that this has been happening only since I upgraded to 6.1.2. The observed frequency seems to be compatible with the hypothesis that the bogus values occur each time another 4GB has been written to the tape. Chris Thompson University of Cambridge Computing Service, Email: cet1(a)ucs.cam.ac.uk New Museums Site, Cambridge CB2 3QH, Phone: +44 1223 334715 United Kingdom.

1 0

Update to a previous email
by Art Hebert 04 Jan '02

04 Jan '02

I posted this recently in response to somebody trying to sell a used netapp: Its more valuable to part it out than sell it. It cost more to have maintenance on it than to buy it. You can buy a terabyte of emc gear for $5K now. Our Netapp salesrep took offense as he thought I was bashing Netapp. I love Netapp, I've been using them for a long time. Its just that used netapps aren't worth it if you don't have a license, the parts are therefore worth more than the sum and at auction you can pick them and emc gear up really cheap. So do I am back in good graces with Netapp, pleaseeeeeee :-) Art Hebert Email: art(a)arzoon.com <mailto:art@arzoon.com> Arzoon IT Dept. Phone: 650 995-6125 Mobile: 650 444-8667

2 1

Replacing a FC Drive Controller Card...
by Dale Gass 04 Jan '02

04 Jan '02

Hey, got a couple of NetApp questions; they're a little low-level, and undoubtedly unsupported and warranty-voiding, etc., but any advice would be appreciated... (We don't have a support contract, due to the costs.) After moving a network appliance, it woke up kind of grumpy. One of the drives (Seagate Cheetah 18G FC-AL) didn't respond at all. Normally, when powering on a shelf, the green light typically goes off, blinks a bit, and then comes on solid. With this drive, the light just stays on solid, from the instant the switch is flipped on for the shelf. Pretty much non-responsive. So the unit attempts a rebuild. However, during the rebuild, unrecoverable media errors are encountered on another drive in the raid set. Sigh... "File system may be scrambled." So... With the netapp off, I try another drive in the non responsive shelf's slot. It's status light does the normal thing, so the shelf is okay, something is wrong with the drive or controller. Given the fact the light doesn't blink at all upon initialization, I suspect the controller. Since I had another spare drive, of exactly the same make/model, I tried swapping the controller card with the non-responsive unit. Doing so, then powering on the shelf, gives the normal status light sequence (blinking, then solid). A good sign so far. Then... Powering on the netapp: it says the raid set is still missing a drive, and shows the drive with the new controller, as "uninitialized", and assigns it to a hot spare, and then tries the rebuild again (which fails on the media errors on the other drive...) So. I'm guessing the NetApp uses the drive's serial # (which is on the controller card, not the drive, I presume) to keep track of the drive's function. I guess my three questions are as follows: 1. Is there any way to tell the NetApp that a drive's serial # has changed? (Where is this low-level raid configuration data stored? In NVRAM I assume? I looked around the files in /vol/vol0/etc, but nothing looked appropriate.) 2. Does the fact that the drive was flagged as a hot spare actually cause anything to be written to the drive, or is it just noted as such in the NetApp's configuration? (Also, since a rebuild was attempted, does that mean my data was overwritten? I guess since it was a rebuild, any data that was successfully rebuilt before the media errors, should be the same as was on the drive before, right? Or not...? The drive #was 26. There was another hot spare at #24; #26 is listed first in the host spare list on boot; would the lower-numbered or listed-first hot spare tend to be used? The filer didn't indicate *which* hot spare it was starting to use in the rebuild.) 3. Each time that the filer starts up, it claims to recover different blocks on the medium error, and claim they'll be reassigned. Is there any way to force the retries on this higher, so it will be more determined at recovery on the medium errors? Since different sectors are successfully read different times, one would think that if it did the reassignment each time, it would eventually recover the drive (maybe). Dual failures really suck on raid. I'm hoping that there's a way to bring this back to life. (We do have the data on tape; but due to a number of circumstances I won't go into here, restoring it would be *very* laborious, so I'm hoping for a bit more of a creative solution :-) Thanks, all... -dale

2 4

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