From sunmanager at kiasoft.com Tue Sep 11 07:24:20 2001 From: sunmanager at kiasoft.com (Jeff Painter) Date: Tue, 11 Sep 2001 11:24:20 GMT Subject: SUMMARY: Upgrade solaris 2.6 to 8, veritas problems Message-ID: <20010911112420.28347.qmail@kiasoft.com> Thanks to everyone that wrote to me on this one. Fred Hall Bernard.L.duBreuil Julie L Baumler, SCSA Andy Cordova Dave Foster Changa Anderson but the last response I got was the simplest, made the most sense and worked for me from Changa Anderson. my initial problem was that I wanted to upgrade my solaris system only (E3500 with solaris 2.6, veritas 3.0.4) to solaris 8 without any other modifications. the simplest means was to execute from the veritas 3.0.4 cd in scripts/upgrade_start and upgrade_finish after the operating system has been put back in. my primary boot drive was not in a veritas volume so the only other step i had to do was to comment out my volumes in /etc/vfstab until after i had executed upgrade_finish. then one last reboot and the volumes came back online. From Hans.Hedlund at enskilda.se Tue Sep 11 07:59:10 2001 From: Hans.Hedlund at enskilda.se (Hans.Hedlund at enskilda.se) Date: Tue, 11 Sep 2001 13:59:10 +0200 Subject: SUMMARY: LVD SCSI on Sun? Message-ID: Original question: > Does Sun have any LVD SCSI PCI adapter? I can only find Differential (X6541A) and Single-Ended (X6540A, X5010A) adapters... Sun seems to want everyone to move to fiber channel, so they don't provide any LVD adapters. See usenet discussion: http://groups.google.com/groups?hl=de&safe=off&th=dc2d8ce8c6581dfa,4 Several people recommends Antares P-00068/69 adapters (www.antares.com), which I think I'll go for myself and hope that my combo of IBM LTO 3583 library and E420R machine works with that adapter (and that my Sun support contract doesn't get voided...) Thanks to everyone for your replies, I just love this list! Regards, Hans ************************************************************************ Confidentiality Notice The content of this e-mail is intended only for the confidential use of the person(s) to whom it is addressed. If the reader of this message is not such a person, you are hereby notified that you have received this communication in error and that reading it, copying it, or in any way disseminating its content to any other person, is strictly prohibited. If you have received this message in error, please notify the author by replying to his e-mail immediately. ************************************************************************ -------------- next part -------------- An HTML attachment was scrubbed... URL: http://www.sunmanagers.org/pipermail/summaries/attachments/20010911/88621636/attachment.html From bowens at eastman.com Tue Sep 11 10:24:33 2001 From: bowens at eastman.com (Owens, Blaine C) Date: Tue, 11 Sep 2001 10:24:33 -0400 Subject: SUMMARY: Eliminate ssh password promp Message-ID: <8FEB1E00D2DAD111A4E00000F8CD1DBC0E106EE5@ntmail20.emn.com> My face has got to be red on this - turned out to be dumb mistake on my part. I had run ssh-keygen on the local machine twice. The second time I ran it created "id_dsa_1024_b" and "id_dsa_1024_b.pub" respectively. It occurred to me that I had failed to change the authentication and identification files to specify my "b" keys instead of my "a" keys. Making this simple change on the local resolved the problem. Anyway, I want to thank ed.rolison at power.alstom.com Bernhard Sadlowski [sadlowsk at mathematik.uni-bielefeld.de] Al Hopper [al at logical-approach.com] for their prompt responses. The $HOME/.shosts file is not needed. Blaine Owens Eastman Chemical Company Phone - (423)-229-3579 Cell Phone - (423)-817-0704 Fax - (423)-229-1188 bowens at eastman.com _______________________________________________ sunmanagers mailing list sunmanagers at sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers From lists at internyc.net Tue Sep 11 13:06:10 2001 From: lists at internyc.net (lists at internyc.net) Date: Tue, 11 Sep 2001 13:06:10 -0400 (EDT) Subject: SUMMARY: Disk Mirroring, root/swap on Netra X1 In-Reply-To: <3B9DA5A8.5C063EA0@internyc.net> Message-ID: Thank you all for a lot of good info: Paul Adair Pankaj Anand Simon Greenland Sun Managers List Mark Lewis Srinivas_Arella Richard Cove Dietsch, Nathan Olivier Masse henrik huhtinen 1. The answer to my first question about mirroring swap was uniform by everyone. Mirror swap so in case that I loose one hard drive the machine will keep on running. Also its easier to maintain the raid and faster. 2. The issue of booting a damaged raid is still somewhat unclear but as soon as I set it up I'll test it and have the right idea. According to some it will boot fine and run (which I doubt). And according to others a failed raid will not be a functional boot device. Henrik mentioned that I should be able to boot into a single user mode with failed raid and fix the raid or diassembel it and then continiue into multi user mode. Also couple of people sugested using the newest Solaris distribution and patches for Netra network cards. thanks fil On Tue, 11 Sep 2001, fil wrote: > Hello, > > I am getting my first Netra X1. I am kind of scared reading about some > quirks with JumpStaring it and missing drivers, oh well... I am planing > of having 2 hard drives in an X1 and want to mirror them. I will use > DiskSuite for that and have done it before without to much troubles on a > regular Sparc Servers (SCSI Drivers). Now here are my 2 main concerns. > > 1. Do I need to mirror a swap partition and how good/bad of an idea is > it and what would be an advantage of it? > 2. If my root partition is mirrored and one of the drives fails can I > still use the other one to boot, or is it hosed because of the drive > failures, messed up fstab info ......? > > > Thanks > fil > _______________________________________________ > sunmanagers mailing list > sunmanagers at sunmanagers.org > http://www.sunmanagers.org/mailman/listinfo/sunmanagers > _______________________________________________ sunmanagers mailing list sunmanagers at sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers From drew at phg.mc.vanderbilt.edu Tue Sep 11 13:31:24 2001 From: drew at phg.mc.vanderbilt.edu (Drew Raines) Date: Tue, 11 Sep 2001 12:31:24 -0500 Subject: SUMMARY: U450 hard disk removal In-Reply-To: <20010910192330.A18458@phg.mc.vanderbilt.edu> References: <20010910192330.A18458@phg.mc.vanderbilt.edu> Message-ID: <20010911123124.M19317@phg.mc.vanderbilt.edu> Most thought it is alright to pull the drive, with the LED on, after unmounting the affected partitions. Now I just need to figure out how to determine which drive I'm looking for out of the 18 in the cabinet. Thanks to: Pankaj Anand | sham sunder | przemol | Amit Batra | Matthew Stier ---------------------------------------------------------------------------- * Drew Raines : > > I need to remove a hot-swappable disk from a U450 running Solaris 7. Do I > need to unconfigure it somehow, or just yank it out and do drvconfig/disks > to rebuild the device info? > > I saw some info on this in the archives concerning Sol 8, but the commands > were esoteric to that version. -- Drew _______________________________________________ sunmanagers mailing list sunmanagers at sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers From drew at phg.mc.vanderbilt.edu Tue Sep 11 14:25:18 2001 From: drew at phg.mc.vanderbilt.edu (Drew Raines) Date: Tue, 11 Sep 2001 13:25:18 -0500 Subject: SUMMARY: U450 hard disk removal (part 2) In-Reply-To: <20010911123124.M19317@phg.mc.vanderbilt.edu> References: <20010910192330.A18458@phg.mc.vanderbilt.edu> <20010911123124.M19317@phg.mc.vanderbilt.edu> Message-ID: <20010911132517.A20021@phg.mc.vanderbilt.edu> * Drew Raines : > > Now I just need to figure out how to determine which drive I'm looking > for out of the 18 in the cabinet. Inspired by all similar suggestions, I figured this out by running format/analyze on the drive and doing a read test. This created a light show on the appropriate drive bay. It was very entertaining. Thanks to Tim Evans, Michael Horton, and Chris Tubutis who all were passengers on this train of thought. Consequently, this confirmed Matthew Stier's listing of the organization of the drives in the cab. 00 = c0t0d0s0 01 = c0t1d0s0 02 = c0t2d0s0 <----- 03 = c0t3d0s0 04 = c4t0d0s0 ... -- Drew _______________________________________________ sunmanagers mailing list sunmanagers at sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers From PZXKYS at atdva3.atd.gmeds.com Tue Sep 11 15:20:27 2001 From: PZXKYS at atdva3.atd.gmeds.com (Martin Meadows) Date: Tue, 11 Sep 2001 14:20:27 -0500 Subject: summary: question abt. rst0 -rw-r--r-- vs. rst1 crw-rw-rw Message-ID: <010911142027.50e76@atdva3.atd.gmeds.com> Thanks to the following folks: Bertrand: Bertrand_Hutin at notes.amdahl.com richard: rbutler at ibc.rm.cnr.it mark: hargrme at wisdom.maf.nasa.gov mystery guru: Stuart.Little at leotel.co.uk steve: steve at ttyl.com Here's the original question: >We're having a problem writing to a tape drive as a non-root user. >I've found that the /dev/rst0 permission scheme is -rw-r--r-- and >it is crw-rw-rw on /dev/rst1. This is a sunos 4.1.3_u1 v.B system. > >1) what is the "c" on /dev/rst1. looks like most of the devices > in /dev have this (except for rst0) >2) if I need the "c" on /def/rst0, how do I get it? >3) Is it okay to chmod 666 /dev/rst0? >4) any idea why it has this oddball permission scheme? Summary: rst0 had been fouled up ... not sure how. I corrected the problem by removing it & running /dev/MAKEDEV st0 And here are the responses: ----------------------------------------------------------------------------- The minor number cannot be selected randomly! try one less than rst1 for a start, otherwise try MAKEDEV... I've forgotten all the old Solaris commands like this, but they will surely do no harm. Delete rmt0 before you start > >The c means it is a character device. To create one, you'll need to use >mknod. If you ls -l rst1, you'll find that instead of a file size, you've >got the major and minor device numbers. At a guess, use the same major, and >minor -1 for rst0 that you find on rst1 ( after you've deleted rst0 first ). > >Once you've successfully created rst0 ( if, of course, you do have 2 tape >drives ), then you can change the permissions to whatever you want. Using >the 'oddball' 666 permissions will, of course, allow anyone to overwrite >your tape! > >man ls and mknod will give you some basic info. > ----------------------------------------------------------------------------- rst0 was probably created as a file when someone tried to do a tar to a device that didn't exist. The 'c' implies that the device is a character device, as opposed to a block device (b prefix) or ordinary file. Assuming you have a tape device available check the kernel configuration file in /usr/sys//conf to ensure support is in kernel for st0, delete/move the existing /dev/rst0 and run './MAKEDEV st0' in /dev. ----------------------------------------------------------------------------- It means it's a character special device and it should be there for rst0 also. Do you mean that root can write to the tape? It looks to me as though /dev/rst0 has somehow been changed into a file by mistake - root would be able to write to the file, but not to the actual tape drive! > 2) if I need the "c" on /def/rst0, how do I get it? Easiest way is with mknod - but you need to know the major and minor block numbers for the device. You can maybe figure them out from rst1, rst2 etc and also nrst0, nrst1 the corresponding no-rewind devices. On my system 4.1.1 these are: rst0 18 0 nrst0 18 4 rst1 18 1 nrst1 18 5 rst2 18 2 nrst2 18 6 ... so the command would be: (in /dev) mknod rst0 c 18 0 (and maybe also for nrst0) > 3) Is it okay to chmod 666 /dev/rst0? Yes (AFAIK), but if rst0 is a file you are just giving them permission to write to it. Create with mknod and then chmod to 666. > > 4) any idea why it has this oddball permission scheme? > /dev/rst0 is either a file or a symbolic link. Do: ls -l "/dev/rst0" and see what you get. I appears that /dev/rst0 got overwritten and is now a tar file. I should have the "c" as part of the permissions if it's not a symbolic link. If it is a file, do the following: You can remove /dev/rst0 and do the following to recreate it: cd /dev MAKEDEV ----------------------------------------------------------------------------- c means character device , correct for a tape drive. /dev/rst0 is a file, not a tape! in /dev there all entries should be c or b devices, not files. The c means it is a character device. To create one, you'll need to use mknod. If you ls -l rst1, you'll find that instead of a file size, you've got the major and minor device numbers. At a guess, use the same major, and minor -1 for rst0 that you find on rst1 ( after you've deleted rst0 first ). Once you've successfully created rst0 ( if, of course, you do have 2 tape drives ), then you can change the permissions to whatever you want. Using the 'oddball' 666 permissions will, of course, allow anyone to overwrite your tape! man ls and mknod will give you some basic info. ----------------------------------------------------------------------------- (response to my question about using mknod command vs. makedev) The Makedev script will only do the same thing. So, you should be ok now. Is this a new tape device that you've just installed? ----------------------------------------------------------------------------- _______________________________________________ sunmanagers mailing list sunmanagers at sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers From jeff at cjsa.com Tue Sep 11 18:11:43 2001 From: jeff at cjsa.com (Jeffery Small) Date: Tue, 11 Sep 2001 15:11:43 -0700 Subject: SUMMARY: Problem with ufsrestore Message-ID: <20010911151143.H1006@cjsa.com> [I took a few days to send this summary as I was waiting for some confirming information from Sun Support.] ****************************************************************************** ORIGINAL QUESTION: ****************************************************************************** I have a DAT-2 tape drive on my SPARCstation-60 which is running Solaris 8. I have five different disk partitions and I have generated a level-0 dump using ufsdump and the non-rewinding device. These five sequential dumps span over three DAT tapes. I wanted to grab something off the tape from the the forth partition using a custom script that basically issue the commands: mt -f /dev/rmt/0c rewind ufsrestore isf 4 /dev/rmt/0c With the first tape in the set installed it begins searching for the 4th dump file. When it gets to the end of the tape it displays the message: Mount volume 2 then enter volume name (default /dev/rmt/0c) I insert the 2nd tape and then have tried the following things: * enter: * enter: /dev/rmt/0c * enter: 2 Regardless of what I do, ufsrestore immediately asks for volume 3, then 4, then 5, and so on. It never makes any attempt to read any subsequent tape. What is the proper response to give at this point to get ufsrestore to continue properly. Thanks for your help. ****************************************************************************** ****************************************************************************** SUMMARY: ****************************************************************************** My confusion was the result of making an assumption that the 's' skip option of ufsrestore would be smart enough to keep track of what it was doing as it advanced beyond the end of the current tape and the next one. Apparently that is not true. To extract from a given dump file, you must start by loading the appropriate tape where that dump begins and then advance an appropriate amount to the start of the target dump file. I will give detailed instructions below which will hopefully be useful to others who may have similar questions. This example should highlight the procedure. On my machine I dumped the following five partitions: / /u /v /opt /var onto a series of three DAT tapes using the non-rewinding device /dev/rmt/0cn. These dumps end up located as followed: Tape #1: / /u /v (beginning) Tape #2: /v (conclusion) /opt (beginning) Tape #3: /opt (conclusion) /var Thus, to access a particular partition, first do "mt -f /dev/rmt/0c rewind" to insure the tape is rewound, and then: Partition Load Tape Command ------------- ------------- ---------------------------------- / #1 ufsrestore isf 1 /dev/rmt/0c /u #1 ufsrestore isf 2 /dev/rmt/0c /v #1 ufsrestore isf 3 /dev/rmt/0c /opt #2 ufsrestore isf 2 /dev/rmt/0c /var #3 ufsrestore isf 2 /dev/rmt/0c Note that if you load tapes #2 or #3 and tried: ufsrestore isf 1 /dev/rmt/0c ufsrestore will appropriately complain: "This is not volume 1 of the dump" since these tapes actually start in the middle of a dump file initiated on the previous tape. Justin Stringfellow of Sun UK sent some notes from the SunSolve database. However, when I attempted to repeat the procedure for reading the tapes in "reverse" order as described below, I discovered that you cannot perform tape drive operations in a second window when the original ufsrestore is running as you get "/dev/rmt/0n: Device busy" errors. I confirmed that this is the case with Sun Support and that the instructions appear to be incorrect. Here is how I went about extracting files from a dump in the middle of the tape set that spans tapes. Let's say we want something from the /opt partition which starts on tape #2 and finishes on tape #3. 1: Insert tape #2 into the drive. 2: Make sure the tape is rewound with: mt -f /dev/rmt/0c rewind 3: Access the /opt interactive index: ufsrestore isf 2 /dev/rmt/0c 4: Select the files to be extracted using the normal ufsrestore commands. 5: When you specify "extract", just start with the current volume, which for this dump is "1". (It is actually "tape" #2, but it is "volume" #1 for the /opt partition dump.) Just search the tapes sequentially rather than in reverse order. The dialog is as follows: ufsrestore > extract You have not read any volumes yet. Unless you know which volume your file(s) are on you should start with the last volume and work towards the first. Specify next volume #: 1 6: When ufsrestore reaches the end of volume 1, eject the tape and insert tape #3 which is "volume" #2 for this dump. The dialog is: You have read volumes: 1 Specify next volume #: 2 Mount volume 2 then enter volume name (default: /dev/rmt/0c) 7: When ufsrestore finishes reading the /opt dump, it concludes with the normal messages: set owner/mode for '.'? [yn] y ufsrestore > quit 8: The tape will rewind (since we didn't use the non-rewinding device). Eject the tape and you're done. I have a set of custom scripts that manage all phases of dump and restore. With this new understanding I will now have to enhance the dump script to parse the output from ufsdump to keep track of and log on which tape each partition dump file begins. Then the restore script will be able to automatically prompt for the correct tape to load in order to access the proper dump. My sincere thanks to the following people who provided quick responses to my question. -- Rasal Kumarage Jay Lessert Buddy Lumpkin Justin Stringfellow Vasilis Iliadis Paul Frederiksen Paul Wilkinson Govindaraju P All of the responses are included below. Regards, -- Jeff C. Jeffery Small CJSA LLC (206) 232-3338 jeff at cjsa.com 7000 E Mercer Way, Mercer Island, WA 98040 ****************************************************************************** ------------------------------------------------------------------------------ ANSWERS RECEIVED: ------------------------------------------------------------------------------ Date: Wed, 5 Sep 2001 18:20:05 GMT From: cypher at punk.net Subject: Re: Problem with ufsrestore Doesn't the output of the ufsdumps indicate (prompting) when a tape was put in during a given dump? If partition 3 dumps on tape 2 and onto tape 3, wouldn't partition 4 be the second file on tape 3? Parsing the ufs output could create the labels you need. ------------------------------------------------------------------------------ Date: Thu, 6 Sep 2001 23:36:10 GMT From: Rasal RKU23 Kumarage Subject: Re: Problem with ufsrestore Hi, Probabaly you can start straight from the tape having the 4th dump & using mt & ufsrestore you can recover the file system. Eg: assume your 4th partition is on 2nd tape & it is the second file. Mount the second tape mt -f /dev/rmt/0cn rewind mt -f /dev/rmt/0cn fsf 2 cd ufsrestore -rvf /dev/rmt/0cn ( You may use ufsrestore -tvf /dev/rmt/0cn to read the tape first) Rasal ------------------------------------------------------------------------------ Date: Thu, 6 Sep 2001 16:48:49 -0700 From: Jay Lessert Subject: Re: Problem with ufsrestore I guess I would be amazed if this worked. :-) I think your script is going to have to keep track of which dumps begin on which file of which tape, and then start with *that* tape (with the skip number !=4 in this case). -- Jay Lessert jay_lessert at accelerant.net Accelerant Networks Inc. (voice)1.503.439.3461 Beaverton OR, USA (fax)1.503.466-9472 ------------------------------------------------------------------------------ Date: Thu, 6 Sep 2001 20:46:13 -0700 From: "Buddy Lumpkin" Subject: RE: Problem with ufsrestore Your trying to restore with interactive options ... don't do that! Try this: Let's say the fourth filesystem backed up is /export/home, and the file you want to restore is everything under /export/home/users/foobar. Your restore command will be: ufsrestore xvfs /dev/rmt/0c 4 ./users/foobar When it asks you which tape, just enter tape #1. If that doesn't work, enter #2, etc.. If you want to predict which tape your files will be on, you need to know how big the backups are. One thing to note is that you can fill four tapes with four backups and only the dumps that span from one tape to the next would have a tape #2, if it spans another tape, then there will be a tape #3 and so on... Also, the no-rewind specifier i.e. /dev/rmt/0cn is only needed during backups to keep the tape from rewinding between backups. Hope this helps, --Buddy ------------------------------------------------------------------------------ Date: Fri, 7 Sep 2001 08:26:06 +0100 (BST) From: Justin Stringfellow Subject: Re: Problem with ufsrestore How to restore dumps that span more than one tape and where there are also multiple dumps on the tapes. For example, using a DAT Autoloader in sequential mode, you have backed up 4 filesystems on 2 tapes as follows: Tape#1 ====== Dump#1 Dump#2 Dump#3 part 1 of 2 ------------------------- ------------------------- -------------------- | Dump | Dump data | | Dump | Dump data | | Dump | Dump data | EOT |Header| | |Header| | |Header| | ------------------------- ------------------------- -------------------- Tape#2 ====== Dump#3 part 2 of 2 Dump#4 -------------------- ------------------------- | Dump | Dump data | | Dump | Dump data | |Header| | |Header| | -------------------- ------------------------- To restore Dump#3, you should: 1) Manually load Tape#1 into the Autoloader 2) Forward the tape to the start of the dump as follows: # mt -f /dev/rmt/0n rewind # mt -f /dev/rmt/0n fsf 2 3) Use the ufsrestore command as follows: # ufsrestore -ivf /dev/rmt/0n ... ufsrestore> add . ... ufsrestore> extract Extract requested files You have not read any volumes yet. Unless you know which volume your file(s) are on you should start with the last volume and work towards the first. Specify next volume #: 4) At this point enter 2, as this dump spans 2 tapes, then manually eject tape#1 and load Tape#2 into the Autoloader. ... Mount volume 1 then enter volume name (default: /dev/rmt/0n) 5) At this point, you should manually , and load Tape#1. 6) To get to the correct place on the tape, you should now go to another window/login on your system position the tape at the start of Dump#3 with: # mt -f /dev/rmt/0n fsf 2 7) Finally, press RETURN to the prompt from ufsrestore to continue the restore. NOTE1: If you position the tape incorrectly, ufsrestore will detect this, and ask you to select another tape. NOTE2: To restore Dump#4, then use must load Tape#2 and foward the tape by one position. You *cannot* load Tape#1, and expect "ufsrestore -ivfs /dev/rmt/0n 3" to autoload Tape#2, as there is no autoloader/jukebox support in Solaris. -- ______ /_____/\ Justin Stringfellow /____ \\ \ Support Engineer /_____\ \\ / Sun Microsystems /_____/ \/ / / Guillemont Park /_____/ / \//\ \_____\//\ / / Tel. : +44 (0)870 6003222 \_____/ / /\ / \_____/ \\ \ \_____\ \\ \_____\/ ------------------------------------------------------------------------------ Date: Fri, 7 Sep 2001 06:59:24 GMT From: vasilis iliadis Subject: RE: Problem with ufsrestore Try this mt -f /dev/rmt/0c rewind ufsrestore xvfs /dev/rmt/0c 4 Regards Vasilis Iliadis Software Engineer Dienekis Information systems SA 446 Mesogion & Marouli Phone: +301 60 17 382 Fax: +301 60 10 690 e-mail: vasilis.iliadis at dienekis.gr ------------------------------------------------------------------------------ Date: Fri, 7 Sep 2001 08:42:22 -0400 From: "Frederiksen, Paul" Subject: RE: Problem with ufsrestore maybe just /dev/rmt/0 ? ------------------------------------------------------------------------------ Date: Fri, 07 Sep 2001 21:07:38 +0200 From: Paul Wilkinson Subject: Re: Problem with ufsrestore I'm sure you have the answer by now, but just in case..... I think the man page is inaccurate in some respects, if you want the 4th file you need to skip 3 which will position you at the beginning of the 4th file. Have you tried forward spacing manually, them restoring. mt -f /dev/rmt/0n fsf 3; ufsrestore ivf /dev/rmt/0n Paul. ------------------------------------------------------------------------------ Date: Sat, 8 Sep 2001 10:11:32 GMT From: "Govindaraju P (OCS-DELRO-AVS)" Subject: Ufsrestore Hi jeff, I too had same problem with DLT4000 drives. It keeps asking volume name.It clicked for me, you can try this option. Load recommended patches for sol8. Rewind the tape without non rewind device file Mount with non rewind device use non rewind file for ufsrestore. It is advisable that you can rewind both the volumes before use then try step 2 and 3. 1.mt -f /dev/rmt/0c rewind 2.mt -f /dev/rmt/ocn stat 3.ufsrestore isf 4 /dev/rmt/0cn Regards, Govind. ------------------------------------------------------------------------------ _______________________________________________ sunmanagers mailing list sunmanagers at sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers From brampling at radiant.net Wed Sep 12 12:16:42 2001 From: brampling at radiant.net (Blair Rampling) Date: Wed, 12 Sep 2001 09:16:42 -0700 Subject: SUMMARY: Automated package installation In-Reply-To: Message-ID: I received a whole lot of answers, several of which recommended the following: Create an "admin" file containing: instance=unique partial=nocheck runlevel=nocheck idepend=quit rdepend=quit space=quit setuid=nocheck conflict=nocheck action=nocheck basedir=default Use pkgadd and specify the admin file using -a This causes the packages not to prompt on these errors as below. -----Original Message----- From: sunmanagers-admin at sunmanagers.org [mailto:sunmanagers-admin at sunmanagers.org]On Behalf Of Blair Rampling Sent: Tuesday, September 11, 2001 8:52 AM To: sunmanagers at sunmanagers.org Subject: Automated package installation How do I automate the installation of a package that prompts me with this: This package contains scripts which will be executed with super-user permission during the process of installing this package. Do you want to continue with the installation of [y,n,?] pkgask doesn't work. Thanks! - -- Blair Rampling Senior Systems Administrator Radiant Communications _______________________________________________ sunmanagers mailing list sunmanagers at sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers _______________________________________________ sunmanagers mailing list sunmanagers at sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers From maccy at maccomms.co.uk Wed Sep 12 15:25:38 2001 From: maccy at maccomms.co.uk (Maccy) Date: Wed, 12 Sep 2001 20:25:38 +0100 Subject: SUMMARY: forms and checklists Message-ID: <05f301c13bc2$af0948c0$0a117ad5@mahabir> My thanks go to :- a) Bara Zani b) Kevin Inscoe Suggestions were (both very useful) :- a) look at www.sun.com/bigadmin they have a checklist for upgrade and install . b) http://inscoe.org/genorders if that helps. Best regards, Mark Mahabir _______________________________________________ sunmanagers mailing list sunmanagers at sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers From maccy at maccomms.co.uk Wed Sep 12 15:37:45 2001 From: maccy at maccomms.co.uk (Maccy) Date: Wed, 12 Sep 2001 20:37:45 +0100 Subject: SUMMARY : shell script for performance testing Message-ID: <05f501c13bc2$b0c550a0$0a117ad5@mahabir> Hi all, Great response to this...... I am indebted to the following (thanks very much guys) :- Joe Fletcher Kurt Werth Bertrand Hutin Ray McCaffity Carsten B Knutsen David Glass Amindra Mahto Olivier Masse Dave Baker Mike Kiernan Johan Hartzenberg Tim Chipman John Leadeham Jay Lessert Suggestions included :- (I went for this one). Try this one that should be able to run unmodified on anything with a Bourne-like shell: #!/bin/sh while : ; do : ; done Other suggestions and relevant websites :- gzipping a large file and loop it www.specbench.org http://setiathome.ssl.berkeley.edu http://www.bitmover.com/lmbench/ http://www.distributed.net/ ----- #!/bin/ksh while true; do wc -c /dev/mem & wc -c /dev/mem & wc -c /dev/mem & wc -c /dev/mem & sleep 10 & done ----- #!/bin/ksh while true do ls -al date done ----- while true do ksh -c "sleep 1" & done ----- #!/usr/bin/ksh while [ x = x ] ; do x=1; done Thanks again all Best wishes Mark Mahabir _______________________________________________ sunmanagers mailing list sunmanagers at sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers From george.done at dutchtone.nl Wed Sep 5 06:44:37 2001 From: george.done at dutchtone.nl (george.done at dutchtone.nl) Date: Wed, 5 Sep 2001 12:44:37 +0200 Subject: Summary: RE: How to disable round robin paging Message-ID: <3F24D8FD4C45D41183CE0008C71E67A502930217@DTEXC03> I've got a lot of interesting messages to the question below. The most important ones were that: 1. There is a solution to this as suggested by Lieven Marchand and this was to change the value of swap_maxcontig from the default 1 Mb to 2GB and have as a secondary effect round robin paging disabled. However, he also pointed out that this is highly experimental and the ones willing to try this on their production systems are on their own. I was not willing to try hence I still have no feedback on this solution. 2. The second (indirect) answer is that applications tend to reserve upfront large quantities of memory upfront from which often a half or even one third actually use. So, the fact that swap apear to be used means that in fact that virtual memory was reserved, not necesarilly used. Of course, because of the above, there will not actually be disk paging activity and subsequently no performance problems. Adrian Cockroft personally pointed me that a lot of memory shortage problems can be solved by just configuring more swap (taking care if 32 bit kernel is used, that none of them is larger than 2GB). In that way, applications which reserves virtual memory will reserve not physical memory but disk swap space. This makes sense. Thanks everybody who bottered to respond ! Best Regards, George DONE > I ran in the following annoying "feature" of Solaris: Round Robin paging. > This are the facts: > 1) the space over 2 GB in a swap device is ignored > 2) if you have more than one swap device, Solaris use them all in a round > robin fashion, several pages on each swap device. > Now this is good if you put each swap device on a separate > volume/metadevice, because spreads the disk load. But if you need 12GB of > swap, that means you should use 12 separate disks to create 6 mirrored > chunks of 2GB. This is way too many disks used for that. > Alternativelly, you could place more than one swap metadevice/volume on > the > same pair of mirrored disks but you will soon notice bad performance > because > of long disk seeks. > Now the question is: > How can I disable this round robin and force Solaris to do paging on only > one volume/metadevice and only when the first one is all used to turn to > the > second one ? The gotcha here is that most of the time I use only 1 GB or > so > of swap for paging and I need to configure 12 GB only to insure I will not > run out of virtual memory in worst case batch run-scenario. The machine is > already filled up with physical memory at full capacity. > > Regards, > George DONE > http://www.sunmanagers.org/mailman/listinfo/sunmanagers > > =========================================================== De verzonden informatie is uitsluitend bestemd voor de geadresseerde natuurlijke persoon of rechtspersoon en bevat mogelijk vertrouwelijke en/of geprivilegeerde gegevens. Met uitzondering van de geadresseerde persoon is het niet toegestaan de informatie openbaar te maken, te kopi?ren, te verspreiden of anderszins actie te ondernemen op basis van de informatie. Indien u de informatie abusievelijk heeft ontvangen, neem dan contact op met de afzender en verwijder de informatie uit alle computers. Dutchtone staat niet in voor de juiste en complete verzending van de informatie, noch is zij aansprakelijk voor de vertraagde ontvangst hiervan. The information transmitted is intended exclusively for the person or entity to which it is addressed and may contain confidential and/or privileged material. Any disclosure, copying, distribution or other action based upon the information by persons or entities other than the intended recipient is prohibited. If you receive this information in error, please contact the sender and delete the material from any and all computers. Dutchtone does not warrant a proper and complete transmission of this information, nor does it accept liability for any delays. =========================================================== _______________________________________________ sunmanagers mailing list sunmanagers at sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers From raghunathl at lucent.com Wed Sep 5 03:08:04 2001 From: raghunathl at lucent.com (RaghuNathL) Date: Wed, 05 Sep 2001 12:38:04 +0530 Subject: SummaryRe: Backup script References: Message-ID: <3B95CF54.8BFFF9A@lucent.com> Hello managers, Iam putting all replies in this mail Hi, I beleive I understood the mail. Here is what i think. Since you are executing the tar from the script itself, it is executed in a subshell. thus the execution of next statement continues. You can create a function and move the tar statement into it. This will hold the execution of Unlock till the function finishes. This could work out. Refer to sh(1) for details on this. Srinivas m: Doug Otto Tue 01:02 Subject: Re: Backup script To: RaghuNathL Modify your script to capture the PID of tar. Once you have it you can use pwait (PID) to watch it. I'm an old csh hack so your syntax may vary.... set PID=`ps -ef | grep tar | grep -v grep | awk '{print $2}'` pwait $PID ...rest of script problem was solved after i put "{" after #1/bin/sh and closed it after done "; }" in the back up script i used also i introduced `date '+%M%H%' to confirm time in the script. RaghuNathL wrote: > Hello Managers, > i use the follwing script to backup clearcase, prblem is tar gives the > control to script at same moment and vob gets unlocked before it's > compltely backedup. > how to make the script to wait until tar is completed? > Any help will be appriciated. > here is the file tags+storage. > /software/nms/net-sim:/net/gangotri/iidcvobs/vobstore/net-sim.vbs > /software/nms/obj:/net/gangotri/iidcvobs/vobstore/obj.vbs > #!/bin/sh > #This script is used for backup od clearcase repository on VOB Server at > > # our facility you need two things here one is a flat file which has > list of all#vobtags and second field as storage path which should be > reachable from backup #host (delimliter is colon). > > # export the SHELL that we are going to use > SHELL=/bin/sh > export SHELL > > # export the correct PATH so that all the required binaries can be found > > PATH=$PATH:/usr/local/bin > export PATH > > # These are the valid statuses which save reports on completion of the > backup > statuses=" failed. abandoned. succeeded. completed savetime= " > > # Perform the PRECMD (Lock VOB) > for _Vobtags in `cat /var/tags+storage` > do > echo "$_Vobtags" > tag=`echo $_Vobtags|cut -f1 -d :` > storage=`echo $_Vobtags|cut -f2 -d :` > sleep 2 > echo "$tag locked" > echo "$storage storage is now getting backedup" > /usr/atria/bin/cleartool setview -exec "/usr/atria/bin/cleartool lock > $tag" backup_view >> /tmp/voblock.log 2>&1 > sleep 2 > /usr/local/bin/tar czvf - $storage |rsh iidcs234 dd of=/dev/rmt/3 > obs=512 > echo "$tag is just now backed up" >> /tmp/backedup > sleep 2 > /usr/atria/bin/cleartool setview -exec "/usr/atria/bin/cleartool unlock > $tag" backup_view >> /tmp/vobunlock.log 2>&1 > sleep 2 > done > > -- > Thanx&Regards > RaghuNathL > Ux-Administrator > CIO-GIO Helpdesk > Lucent-Ins > Ph:5500061 Ex:2048 > "Change in themselves cause changes" -- Thanx&Regards RaghuNathL Ux-Administrator CIO-GIO Helpdesk Lucent-Ins Ph:5500061 Ex:2048 "Change in themselves cause changes" _______________________________________________ sunmanagers mailing list sunmanagers at sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers From compquestion at hotmail.com Wed Sep 12 14:57:09 2001 From: compquestion at hotmail.com (comp question) Date: Wed, 12 Sep 2001 18:57:09 +0000 Subject: SUMMARY:Need help accessing a directory Message-ID: Thanks, as always, to those who replied. Maybe I did not clearly explain my original posting well enough, but most of the answers gave me the basic operations to perform (ie chmod 777 /us/lib/macros, rm -R /usr/lib/macros,etc.) Unfortunately, I tried all of these and many more prior to posting. As of yet I have not gotten a final answer to this problems. The best answer I received was to put the machine in single-user mode and try the operations form there. However, that will not be possible for quite some time as this is a production machine. I will post again once I get this resolved. Thanks MY ORIGINAL POST: I have a directory that appeared on my last reboot and is causing some problems for my backup software. The problem is a directory /usr/lib/macros, which has the permissions: dr-xr-xr-x root root 1 Sep 2 19:07 macros For some reason I cannot access, remove, change, cd, etc into this directory. Before you ask, yes I am root when trying this. I have tried and ls -l on the dir but cannot see it (permission denied), I have tried chmod -f, rm and ln -s /usr/lib/macros /tmp/macros but all not allowed me access the directory. Any ideas, or suggestions? Thanks in advance. _________________________________________________________________ Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp _______________________________________________ sunmanagers mailing list sunmanagers at sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers From petino at hotmail.com Thu Sep 13 03:00:07 2001 From: petino at hotmail.com (Peter Ondruska) Date: Thu, 13 Sep 2001 09:00:07 +0200 Subject: SUMMARY: Bad PBR sig Message-ID: Many thanks go to Lusty Wench . The solution is to omit "filesys rootdisk.s2 all overlap" in jumpstart profile. filesys part should look like this: filesys rootdisk.s0 128 / logging,noatime filesys rootdisk.s1 512 swap filesys rootdisk.s3 1024 /usr logging,noatime filesys rootdisk.s4 1024 /var logging,noatime filesys rootdisk.s5 1024 /opt logging,noatime filesys rootdisk.s6 free /export/home logging,noatime My question was: After a jumpstart installation of Solaris 8_x86 7/01 and reboot I cannot boot the system. I do not even see Device Configuration Assistant. The screen is black and last line shows "Bad PBR sig". There was Solaris installed before on this computer and worked very well. I tried erasing the disk (dd if=/dev/zero of=/dev/rdsk/c0d0p0) but this has no affect. Peter _________________________________________________________________ Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp _______________________________________________ sunmanagers mailing list sunmanagers at sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers From mkiernan at onet.pl Thu Sep 13 07:21:57 2001 From: mkiernan at onet.pl (Mike Kiernan) Date: Thu, 13 Sep 2001 13:21:57 +0200 Subject: SUMMARY(?): Sol8/Disksuite 4.2.1 Raid 5 Data Loss Postmortem Message-ID: <3BA096D5.5220B61A@onet.pl> First question I've sent here and not received a single answer. In my experience this usually means the question is badly phrased or just too boring ;). On either count I apologize :) I still don't understand how/why disksuite got itself into this state. I guess I'll file this in the 'Acts of God' folder. all the best, Mike -- Michael Kiernan Onet.pl S.A. Krakow, Poland http://www.onet.pl/ > Last night we lost a 170GB Disksuite RAID-5 volume. > We were unable to rescue any data and are still > running a recovery. > > E420R running Solaris 8, Disksuite 4.2.1. > > There are 6 disks in the volume, all on controller 1 > (no comments). At the time, metastat reported one > disk as 'Maintenance' then another as 'Last Erred'. > > c1t10d0s0 5082 No Maintenance > c1t11d0s0 5082 No Okay > c1t12d0s0 5082 No Last Erred > > sd24 = c1t10d0s0 -> ../../devices/pci at 1f,4000/scsi at 3,1/sd at a,0:a > sd25 = c1t11d0s0 -> ../../devices/pci at 1f,4000/scsi at 3,1/sd at b,0:a > sd26 = c1t12d0s0 -> ../../devices/pci at 1f,4000/scsi at 3,1/sd at c,0:a > > The first error (in the messages file) does indeed > appear to have occurred on the disk that was marked > Maintenance, and the second error on the Last Erred > disk. The timestamp for the messages is the same. > An error appears a few secs later for the 3rd disk > shown above. > > The problem is that the disk that had actually failed > was the one that was marked 'Last Erred' (ie t12) and > not the one that was marked 'Maintenance' (t10). > On replacing the physically bad disk t12 it was thus > not possible to add t10 back [there is no way to remove > the newly added empty disk from the volume]. metareplacing > t10 would have caused further data corruption. > > I suspect that even if we had managed to recover the > volume there was data corruption [look at the vxfs > errors below]. > > What I want to know is: > > a) Why was the wrong disk marked bad. > b) Why is there data corruption in these 'last erred' scenarios? > c) Is there any way to recover from this situation. > d) bonus question: where does disksuite store raid5 vol > status information [the /etc/md files?] > > Will Summarize. > > Regards, > Mike Kiernan > > [ more data/messages below ] > > 533 $>uname -a > SunOS random 5.8 Generic_108528-06 sun4u sparc > > 532 $>pkginfo -l SUNWmdu > PKGINST: SUNWmdu > NAME: Solstice DiskSuite Commands > CATEGORY: system > ARCH: sparc > VERSION: 4.2.1,REV=1999.11.04.18.29 > BASEDIR: / > VENDOR: Sun Microsystems, Inc. > DESC: Solstice DiskSuite Commands > PSTAMP: 11/04/99-18:31:00 > INSTDATE: Apr 04 2001 10:10 > VSTOCK: 258-6252-11 > HOTLINE: Please contact your local service provider > STATUS: completely installed > FILES: 57 installed pathnames > 11 shared pathnames > 12 directories > 20 executables > 11230 blocks used (approx) > > Did not have the Disksuite 4.2.1 product patch 108693-07, > but this patch contains nothing relevant. > > 534 $>metastat -p | grep d20 > d20 -r c1t2d0s0 c1t3d0s0 c1t4d0s0 c1t10d0s0 c1t11d0s0 c1t12d0s0 -k -i > 32b > > 535 $>metastat > d20: RAID > State: Needs Maintenance > Invoke: metareplace -f d20 c1t10d0s0 > Interlace: 128 blocks > Size: 358355376 blocks > Original device: > Size: 358356480 blocks > Device Start Block Dbase State Hot Spare > c1t2d0s0 5082 No Okay > c1t3d0s0 5082 No Okay > c1t4d0s0 5082 No Okay > c1t10d0s0 5082 No Maintenance > c1t11d0s0 5082 No Okay > c1t12d0s0 5082 No Last Erred > > the disks are: > sd24 = c1t10d0s0 -> ../../devices/pci at 1f,4000/scsi at 3,1/sd at a,0:a > sd25 = c1t11d0s0 -> ../../devices/pci at 1f,4000/scsi at 3,1/sd at b,0:a > sd26 = c1t12d0s0 -> ../../devices/pci at 1f,4000/scsi at 3,1/sd at c,0:a > > /var/adm/messages: > > Sep 11 14:57:51 random scsi: [ID 107833 kern.warning] WARNING: > /pci at 1f,4000/scsi at 3,1/sd at a,0 (sd24): > Sep 11 14:57:51 random SCSI transport failed: reason > 'incomplete': retrying command > Sep 11 14:57:51 random > Sep 11 14:57:51 random scsi: [ID 107833 kern.warning] WARNING: > /pci at 1f,4000/scsi at 3,1/sd at c,0 (sd26): > Sep 11 14:57:51 random SCSI transport failed: reason > 'incomplete': retrying command > Sep 11 14:57:51 random > Sep 11 14:57:52 random scsi: [ID 107833 kern.warning] WARNING: > /pci at 1f,4000/scsi at 3,1/sd at b,0 (sd25): > Sep 11 14:57:52 random SCSI transport failed: reason > 'incomplete': retrying command > Sep 11 14:57:52 random > <-- snip > Sep 11 14:58:58 random scsi: [ID 365881 kern.info] /pci at 1f,4000/scsi at 3,1 > (glm1): > Sep 11 14:58:58 random Cmd (0x3f094030) dump for Target 12 Lun > 0: > Sep 11 14:58:58 random scsi: [ID 365881 kern.info] /pci at 1f,4000/scsi at 3,1 > (glm1): > Sep 11 14:58:58 random cdb=[ 0x28 0x0 0x3 0x9f 0xd9 0xe6 > 0x0 0x0 0x4 0x0 ] > Sep 11 14:58:58 random scsi: [ID 365881 kern.info] /pci at 1f,4000/scsi at 3,1 > (glm1): > Sep 11 14:58:58 random pkt_flags=0x4000 pkt_statistics=0x61 > pkt_state=0x7 > Sep 11 14:58:58 random scsi: [ID 365881 kern.info] /pci at 1f,4000/scsi at 3,1 > (glm1): > Sep 11 14:58:58 random pkt_scbp=0x0 cmd_flags=0x8e1 > Sep 11 14:58:58 random scsi: [ID 365881 kern.info] /pci at 1f,4000/scsi at 3,1 > (glm1): > Sep 11 14:58:58 random Cmd (0x3f095b30) dump for Target 12 Lun > 0: > Sep 11 14:58:58 random scsi: [ID 365881 kern.info] /pci at 1f,4000/scsi at 3,1 > (glm1): > Sep 11 14:58:58 random cdb=[ 0x28 0x0 0x0 0x7b 0x1 0x4a > 0x0 0x0 0x4 0x0 ] > Sep 11 14:58:58 random scsi: [ID 365881 kern.info] /pci at 1f,4000/scsi at 3,1 > (glm1): > Sep 11 14:58:58 random pkt_flags=0x4000 pkt_statistics=0x61 > pkt_state=0x7 > Sep 11 14:58:58 random scsi: [ID 365881 kern.info] /pci at 1f,4000/scsi at 3,1 > (glm1): > Sep 11 14:58:58 random pkt_scbp=0x0 cmd_flags=0x8e1 > Sep 11 14:58:58 random scsi: [ID 365881 kern.info] /pci at 1f,4000/scsi at 3,1 > (glm1): > Sep 11 14:58:58 random Cmd (0x3f36f6c0) dump for Target 12 Lun > 0: > Sep 11 14:58:58 random scsi: [ID 365881 kern.info] /pci at 1f,4000/scsi at 3,1 > (glm1): > Sep 11 14:58:58 random cdb=[ 0x28 0x0 0x0 0xdb 0xd9 0x4e > 0x0 0x0 0x4 0x0 ] > Sep 11 14:58:58 random scsi: [ID 365881 kern.info] /pci at 1f,4000/scsi at 3,1 > (glm1): > Sep 11 14:58:58 random pkt_flags=0x4000 pkt_statistics=0x61 > pkt_state=0x7 > Sep 11 14:58:58 random scsi: [ID 365881 kern.info] /pci at 1f,4000/scsi at 3,1 > (glm1): > Sep 11 14:58:58 random pkt_scbp=0x0 cmd_flags=0x8e1 > Sep 11 14:58:58 random scsi: [ID 107833 kern.warning] WARNING: > /pci at 1f,4000/scsi at 3,1 (glm1): > Sep 11 14:58:58 random Disconnected tagged cmd(s) (3) timeout > for Target 12.0 > Sep 11 14:58:58 random glm: [ID 401478 kern.warning] WARNING: > ID[SUNWpd.glm.cmd_timeout.6018] > Sep 11 14:58:58 random scsi: [ID 107833 kern.warning] WARNING: > /pci at 1f,4000/scsi at 3,1/sd at c,0 (sd26): > Sep 11 14:58:58 random SCSI transport failed: reason 'reset': > retrying command > Sep 11 14:58:58 random > Sep 11 14:58:58 random scsi: [ID 107833 kern.warning] WARNING: > /pci at 1f,4000/scsi at 3,1/sd at c,0 (sd26): > Sep 11 14:58:58 random SCSI transport failed: reason 'timeout': > retrying command > Sep 11 14:58:58 random > Sep 11 14:59:03 random scsi: [ID 107833 kern.warning] WARNING: > /pci at 1f,4000/scsi at 3,1/sd at a,0 (sd24): > Sep 11 14:59:03 random Error for Command: > read(10) Error Level: Fatal > Sep 11 14:59:03 random scsi: [ID 107833 kern.notice] Requested Block: > 57638178 Error Block: 57638178 > Sep 11 14:59:03 random scsi: [ID 107833 kern.notice] Vendor: > SEAGATE Serial Number: 3DE047KQ > Sep 11 14:59:03 random scsi: [ID 107833 kern.notice] Sense Key: Not > Ready > Sep 11 14:59:03 random scsi: [ID 107833 kern.notice] ASC: 0x4 ( unique code 0x4>), ASCQ: 0x1, FRU: 0x2 > Sep 11 14:59:03 random md_raid: [ID 112651 kern.warning] WARNING: md > d20: read error on /dev/dsk/c1t10d0s0 > Sep 11 14:59:04 random proftpd[26721]: [ID 567783 daemon.notice] random > (dial521.walbrzych.dialog.net.pl[217.30.158.13]) - FSep 11 14:59:05 > random proftpd[26726]: [ID 567783 daemon.notice] random > (dial521.walbrzych.dialog.net.pl[217.30.158.13]) - FSep 11 14:59:08 > random scsi: [ID 107833 kern.warning] WARNING: > /pci at 1f,4000/scsi at 3,1/sd at b,0 (sd25): > Sep 11 14:59:08 random Error for Command: > read(10) Error Level: Fatal > Sep 11 14:59:08 random scsi: [ID 107833 kern.notice] Requested Block: > 32651478 Error Block: 32651478 > Sep 11 14:59:08 random scsi: [ID 107833 kern.notice] Vendor: > SEAGATE Serial Number: 3DE046S4 > Sep 11 14:59:08 random scsi: [ID 107833 kern.notice] Sense Key: Not > Ready > Sep 11 14:59:08 random scsi: [ID 107833 kern.notice] ASC: 0x4 ( unique code 0x4>), ASCQ: 0x1, FRU: 0x2 > Sep 11 14:59:08 random md_raid: [ID 371651 kern.warning] WARNING: md > d20: write error on /dev/dsk/c1t11d0s0 > Sep 11 14:59:08 random scsi: [ID 107833 kern.warning] WARNING: > /pci at 1f,4000/scsi at 3,1/sd at a,0 (sd24): > Sep 11 14:59:08 random Error for Command: > read(10) Error Level: Fatal > Sep 11 14:59:08 random scsi: [ID 107833 kern.notice] Requested Block: > 47632810 Error Block: 47632810 > Sep 11 14:59:08 random scsi: [ID 107833 kern.notice] Vendor: > SEAGATE Serial Number: 3DE047KQ > Sep 11 14:59:08 random scsi: [ID 107833 kern.notice] Sense Key: Not > Ready > Sep 11 14:59:08 random scsi: [ID 107833 kern.notice] ASC: 0x4 ( unique code 0x4>), ASCQ: 0x1, FRU: 0x2 > Sep 11 14:59:08 random md_raid: [ID 112651 kern.warning] WARNING: md > d20: read error on /dev/dsk/c1t10d0s0 > Sep 11 14:59:11 random proftpd[395]: [ID 567783 daemon.warning] random - > MaxInstances (50) reached, new connection denied. > Sep 11 14:59:11 random scsi: [ID 107833 kern.warning] WARNING: > /pci at 1f,4000/scsi at 3,1/sd at c,0 (sd26): > Sep 11 14:59:11 random Error for Command: > read(10) Error Level: Fatal > Sep 11 14:59:11 random scsi: [ID 107833 kern.notice] Requested Block: > 8441182 Error Block: 8441182 > Sep 11 14:59:11 random scsi: [ID 107833 kern.notice] Vendor: > SEAGATE Serial Number: 3DE046W4 > Sep 11 14:59:11 random scsi: [ID 107833 kern.notice] Sense Key: Not > Ready > Sep 11 14:59:11 random scsi: [ID 107833 kern.notice] ASC: 0x4 ( unique code 0x4>), ASCQ: 0x1, FRU: 0x2 > Sep 11 14:59:11 random md_raid: [ID 112651 kern.warning] WARNING: md > d20: read error on /dev/dsk/c1t12d0s0 > Sep 11 14:59:13 random proftpd[395]: [ID 567783 daemon.warning] random - > MaxInstances (50) reached, new connection denied. > Sep 11 14:59:13 random scsi: [ID 107833 kern.warning] WARNING: > /pci at 1f,4000/scsi at 3,1/sd at b,0 (sd25): > Sep 11 14:59:13 random Error for Command: > read(10) Error Level: Fatal > Sep 11 14:59:13 random scsi: [ID 107833 kern.notice] Requested Block: > 41253546 Error Block: 41253546 > Sep 11 14:59:13 random scsi: [ID 107833 kern.notice] Vendor: > SEAGATE Serial Number: 3DE046S4 > Sep 11 14:59:13 random scsi: [ID 107833 kern.notice] Sense Key: Not > Ready > Sep 11 14:59:13 random scsi: [ID 107833 kern.notice] ASC: 0x4 ( unique code 0x4>), ASCQ: 0x1, FRU: 0x2 > Sep 11 14:59:13 random scsi: [ID 107833 kern.warning] WARNING: > /pci at 1f,4000/scsi at 3,1/sd at a,0 (sd24): > Sep 11 14:59:13 random Error for Command: > read(10) Error Level: Fatal > > <-- snip > Sep 11 15:01:38 friko02 vxfs: [ID 710715 kern.warning] WARNING: msgcnt 1 > vxfs: mesg 037: vx_metaioerr - /dev/md/dsk/d20 file system meta data > write error > Sep 11 15:01:38 friko02 vxfs: [ID 486312 kern.warning] WARNING: msgcnt 2 > vxfs: mesg 025: vx_wsuper - /dev/md/dsk/d20 file system super-block > update failed > Sep 11 15:01:38 friko02 vxfs: [ID 863296 kern.warning] WARNING: msgcnt 3 > vxfs: mesg 017: vx_delbuf_flush - /fw1 file system inode 3280982 marked > bad > Sep 11 15:01:38 friko02 vxfs: [ID 885974 kern.info] vxfs msgcnt 3 offset > 0x00000000 41ed 5 258 258 > Sep 11 15:01:38 friko02 vxfs: [ID 885974 kern.info] vxfs msgcnt 3 offset > 0x00000010 0 2000 3b9e0a4d 8ccb6 > Sep 11 15:01:38 friko02 vxfs: [ID 885974 kern.info] vxfs msgcnt 3 offset > 0x00000020 3b9e0a4d 80966 3b9e0a4d 80966 > Sep 11 15:01:38 friko02 vxfs: [ID 885974 kern.info] vxfs msgcnt 3 offset > 0x00000030 10000 0 676a2f 0 > Sep 11 15:01:38 friko02 vxfs: [ID 885974 kern.info] vxfs msgcnt 3 offset > 0x00000040 1 1 0 92 > Sep 11 15:01:38 friko02 vxfs: [ID 885974 kern.info] vxfs msgcnt 3 offset > 0x00000050 0 0 0 0 > Sep 11 15:01:38 friko02 vxfs: [ID 885974 kern.info] vxfs msgcnt 3 offset > 0x00000060 141b240 1 0 0 > Sep 11 15:01:38 friko02 vxfs: [ID 885974 kern.info] vxfs msgcnt 3 offset > 0x00000070 0 0 0 0 > Sep 11 15:01:38 friko02 vxfs: [ID 885974 kern.info] vxfs msgcnt 3 offset > 0x00000080 0 0 0 0 > Sep 11 15:01:38 friko02 vxfs: [ID 885974 kern.info] vxfs msgcnt 3 offset > 0x00000090 0 0 0 0 > Sep 11 15:01:38 friko02 vxfs: [ID 885974 kern.info] vxfs msgcnt 3 offset > 0x000000a0 0 0 0 0 > Sep 11 15:01:38 friko02 vxfs: [ID 214594 kern.info] vxfs msgcnt 3 offset > 0x000000b0 0 0 > Sep 11 15:01:38 friko02 vxfs: [ID 486440 kern.warning] WARNING: msgcnt 4 > vxfs: mesg 025: vx_wsuper - /dev/md/dsk/d20 file systSep 11 15:01:38 > friko02 vxfs: [ID 217438 kern.warning] WARNING: msgcnt 5 vxfs: mesg 015: > vx_ilisterr - /fw1 file system can't > Sep 11 15:01:38 friko02 vxfs: [ID 962359 kern.warning] WARNING: msgcnt 6 > vxfs: mesg 031: vx_disable - /dev/md/dsk/d20 file sys > > <---snip > Sep 11 15:01:48 friko02 vxfs: [ID 885974 kern.info] vxfs msgcnt 36128344 > offset 0x00000000 80100 17481e04 191404f0 1c541a4c > Sep 11 15:01:48 friko02 vxfs: [ID 885974 kern.info] vxfs msgcnt 36128344 > offset 0x00000010 9bc1248 1b280000 17341fb4 16c4 > Sep 11 15:01:48 friko02 vxfs: [ID 885974 kern.info] vxfs msgcnt 36128344 > offset 0x00000020 e0c 165418f4 c8003b0 1cb01018 > Sep 11 15:01:48 friko02 vxfs: [ID 885974 kern.info] vxfs msgcnt 36128344 > offset 0x00000030 ff41ce4 1098 d900000 1a281e14 _______________________________________________ sunmanagers mailing list sunmanagers at sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers From rob at garf.nl Thu Sep 13 11:07:53 2001 From: rob at garf.nl (Rob Lyle) Date: Thu, 13 Sep 2001 16:07:53 +0100 Subject: SUMMARY: SPARCstation 5 - Sol 8 on ST39150N Message-ID: <013a01c13c65$dd0e9b50$03e661c3@CONPC002> Thanks to Steve, Will and Peter. You know who you are. My disk was U/S. Simple. Thanks, --Rob. ST39150N > > Hi, > > > > I'm trying to put Solaris 8 onto an aging SPARCstation 5 with a 9GB > Seagate > > disk, in order to build an install server dedicated to the X1 revolution, > > ahem. > > > > After s.l.o.w.l.y. booting from CD and starting the Solaris 8 Webstart > > installer 3.0, the install kernel boots and searches for disks, during > which > > time the ST39140N goes "offline". Error given below: > > > > WARNING: > > /iommu at 0,10000000/sbus at 0,10001000/espdma5 at 0,840000/esp at 5,8800000/sd at 1,0 > > > > probe-scsi sees the disk fine (target 1 at the mo'). > > > > Am I missing a jumper on the disk to stop it spinning down? If not, what > > have I missed? > > > > Thanks, > > > > --Rob. > > > > > > _______________________________________________ > > sunmanagers mailing list > > sunmanagers at sunmanagers.org > > http://www.sunmanagers.org/mailman/listinfo/sunmanagers > > > _______________________________________________ sunmanagers mailing list sunmanagers at sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers From peterv at easics.be Thu Sep 13 11:35:09 2001 From: peterv at easics.be (Peter Vos) Date: Thu, 13 Sep 2001 17:35:09 +0200 Subject: Summary: dd to copy harddisk Message-ID: <3BA0D22D.18AC6722@easics.be> Hello, First of all, thanks to all the responses from (hopefully I don't forget anyone): "Neill, Mark (MBS)" "JULIAN, JOHN C. (AIT)" Hindley Nick Russ Poffenberger todd.a.fiedler at mail.sprint.com Bertrand_Hutin at notes.amdahl.com "Eriksson (London), Christer" Tony Tran "Ricca, Paul" Edward Scown "Caterina Brott" Steve Camp "Wade, Wally" Lars Oliver Bausch "Kruchkoff, A (Alex)" "Pankaj Anand" Tony Walsh Gabel Martin Paul Updike -----original message:----- The disk of one of our machines (Ultra2) has begun making 'chirping' noises, so we will change it with another one. I would like to use dd to copy the existing disk - which contains the OS - with dd to a new disk. Does this also work when using 2 different disks? (The noisy disk is a 2.1G, and will probably be replaced by a bigger one) What other things do I have to keep in mind? (apart from creating the partitions and doing the dd) -----summary:----- Almost everyone advised against using dd, and use ufsdump/restore instead, followed by installboot to make the disk bootable. It's not possible to list everyone's answer, but I will include the most detailed (from "Ricca, Paul" ): INFODOC ID: 16311 SYNOPSIS: How to do a disk-to-disk copy via ufsdump/tar under Solaris 2.X. DETAIL DESCRIPTION: For Solaris 2.X: ________________ Follow the steps below to add a new external/internal disk: [1.] Bring the system down to the ok prompt. # init 0 [2.] Find an available target setting. This command will show what you currently have on your system. # probe-scsi If the disk is on another scsi controller (another card off of an sbus slot) # probe-scsi-all [3.] Attach the new disk with the correct target setting. Run probe-scsi again to make sure the system sees it. If it doesn't, the disk is either not connected properly, has a target conflict, or is defective. Resolve this issue before continuing. In this example, we'll say: T3 original internal drive T1 new/other internal drive where a duplicate copy of the OS will be placed. [4.] Perform a reconfiguration boot. # boot -rv rv -> reconfigure in verbose mode. [5.] Run format and partition the disk. (Here's our example): # format Searching for disks...done AVAILABLE DISK SELECTIONS: 1. c0t1d0 /iommu at 0,10000000/sbus at 0,10001000/espdma at 5,8400000/esp at 5,8800000/sd at 1,0 2. c0t3d0 /iommu at 0,10000000/sbus at 0,10001000/espdma at 5,8400000/esp at 5,8800000/sd at 3,0 Specify disk (enter its number): 1 selecting c0t1d0 [disk formatted] FORMAT MENU: disk - select a disk type - select (define) a disk type partition - select (define) a partition table current - describe the current disk format - format and analyze the disk repair - repair a defective sector label - write label to the disk analyze - surface analysis defect - defect list management backup - search for backup labels verify - read and display labels save - save new disk/partition definitions inquiry - show vendor, product and revision volname - set 8-character volume name quit format> part PARTITION MENU: 0 - change `0' partition 1 - change `1' partition 2 - change `2' partition 3 - change `3' partition 4 - change `4' partition 5 - change `5' partition 6 - change `6' partition 7 - change `7' partition select - select a predefined table modify - modify a predefined partition table name - name the current table print - display the current table label - write partition map and label to the disk quit partition> print Current partition table (original): Total disk cylinders available: 2036 + 2 (reserved cylinders) Part Tag Flag Cylinders Size Blocks 0 root wm 0 - 203 100.41MB (204/0/0) 205632 1 swap wu 204 - 407 100.41MB (204/0/0) 205632 2 backup wm 0 - 2035 1002.09MB (2036/0/0) 2052288 3 unassigned wm 0 0 (0/0/0) 0 4 var wm 408 - 611 100.41MB (204/0/0) 205632 5 unassigned wm 612 - 1018 200.32MB (407/0/0) 410256 6 usr wm 1019 - 2034 500.06MB (1016/0/0) 1024128 7 unassigned wm 0 0 (0/0/0) 0 partition> ****** Modify partitions to suit your needs ****** ****** Do NOT alter partition 2, backup !!! ****** In this example we'll go with the current displayed partition table listed: partition> 0 Part Tag Flag Cylinders Size Blocks 0 unassigned wm 0 - 162 80.23MB (163/0/0) 164304 Enter partition id tag[unassigned]: Enter partition permission flags[wm]: Enter new starting cyl[0]: o `o' is not an integer. Enter new starting cyl[0]: 0 Enter partition size[164304b, 163c, 80.23mb, 0.08gb]: 100.41mb partition> pr Current partition table (unnamed): Total disk cylinders available: 2036 + 2 (reserved cylinders) Part Tag Flag Cylinders Size Blocks 0 unassigned wm 0 - 203 100.41MB (204/0/0) 205632 1 unassigned wu 163 - 423 128.46MB (261/0/0) 263088 2 backup wu 0 - 2035 1002.09MB (2036/0/0) 2052288 3 unassigned wm 0 0 (0/0/0) 0 4 unassigned wm 424 - 749 160.45MB (326/0/0) 328608 5 unassigned wm 750 - 1109 177.19MB (360/0/0) 362880 6 unassigned wm 1110 - 2035 455.77MB (926/0/0) 933408 7 unassigned wm 0 0 (0/0/0) 0 partition> 1 Part Tag Flag Cylinders Size Blocks 1 unassigned wu 163 - 423 128.46MB (261/0/0) 263088 Enter partition id tag[unassigned]: Enter partition permission flags[wu]: Enter new starting cyl[163]: 204 Enter partition size[263088b, 261c, 128.46mb, 0.13gb]: 100.41mb partition> pr Current partition table (unnamed): Total disk cylinders available: 2036 + 2 (reserved cylinders) Part Tag Flag Cylinders Size Blocks 0 unassigned wm 0 - 203 100.41MB (204/0/0) 205632 1 unassigned wu 204 - 407 100.41MB (204/0/0) 205632 2 backup wu 0 - 2035 1002.09MB (2036/0/0) 2052288 3 unassigned wm 0 0 (0/0/0) 0 4 unassigned wm 424 - 749 160.45MB (326/0/0) 328608 5 unassigned wm 750 - 1109 177.19MB (360/0/0) 362880 6 unassigned wm 1110 - 2035 455.77MB (926/0/0) 933408 7 unassigned wm 0 0 (0/0/0) 0 partition> 4 Part Tag Flag Cylinders Size Blocks 4 unassigned wm 424 - 749 160.45MB (326/0/0) 328608 Enter partition id tag[unassigned]: Enter partition permission flags[wm]: Enter new starting cyl[424]: 408 Enter partition size[328608b, 326c, 160.45mb, 0.16gb]: 100.41mb partition> pr Current partition table (unnamed): Total disk cylinders available: 2036 + 2 (reserved cylinders) Part Tag Flag Cylinders Size Blocks 0 unassigned wm 0 - 203 100.41MB (204/0/0) 205632 1 unassigned wu 204 - 407 100.41MB (204/0/0) 205632 2 backup wu 0 - 2035 1002.09MB (2036/0/0) 2052288 3 unassigned wm 0 0 (0/0/0) 0 4 unassigned wm 408 - 611 100.41MB (204/0/0) 205632 5 unassigned wm 750 - 1109 177.19MB (360/0/0) 362880 6 unassigned wm 1110 - 2035 455.77MB (926/0/0) 933408 7 unassigned wm 0 0 (0/0/0) 0 partition> 5 Part Tag Flag Cylinders Size Blocks 5 unassigned wm 750 - 1109 177.19MB (360/0/0) 362880 Enter partition id tag[unassigned]: Enter partition permission flags[wm]: Enter new starting cyl[750]: 612 Enter partition size[362880b, 360c, 177.19mb, 0.17gb]: 177mb partition> pr Current partition table (unnamed): Total disk cylinders available: 2036 + 2 (reserved cylinders) Part Tag Flag Cylinders Size Blocks 0 unassigned wm 0 - 203 100.41MB (204/0/0) 205632 1 unassigned wu 204 - 407 100.41MB (204/0/0) 205632 2 backup wu 0 - 2035 1002.09MB (2036/0/0) 2052288 3 unassigned wm 0 0 (0/0/0) 0 4 unassigned wm 408 - 611 100.41MB (204/0/0) 205632 5 unassigned wm 612 - 971 177.19MB (360/0/0) 362880 6 unassigned wm 1110 - 2035 455.77MB (926/0/0) 933408 7 unassigned wm 0 0 (0/0/0) 0 partition> 6 Part Tag Flag Cylinders Size Blocks 6 unassigned wm 1110 - 2035 455.77MB (926/0/0) 933408 Enter partition id tag[unassigned]: Enter partition permission flags[wm]: Enter new starting cyl[1110]: 972 Enter partition size[933408b, 926c, 455.77mb, 0.45gb]: $ partition> pr Current partition table (unnamed): Total disk cylinders available: 2036 + 2 (reserved cylinders) Part Tag Flag Cylinders Size Blocks 0 unassigned wm 0 - 203 100.41MB (204/0/0) 205632 1 unassigned wu 204 - 407 100.41MB (204/0/0) 205632 2 backup wu 0 - 2035 1002.09MB (2036/0/0) 2052288 3 unassigned wm 0 0 (0/0/0) 0 4 unassigned wm 408 - 611 100.41MB (204/0/0) 205632 5 unassigned wm 612 - 971 177.19MB (360/0/0) 362880 6 unassigned wm 972 - 2035 523.69MB (1064/0/0) 1072512 7 unassigned wm 0 0 (0/0/0) 0 partition> NOTE: You will know for certain that your partitioning is correct if you add all the cylinder values [the values enclosed in ( )], like so, 204+204+204+360+1064=2036 which is the same value for slice 2 or the whole disk (Tag = backup). Now label the disk. This is important as this is what saves the partition table in your VTOC (Virtual Table Of Contents). It's also always recommended to do the labeling part twice to be certain that the VTOC gets saved. partition> label partition> q format> q After partitioning c0t1d0 to be exactly the same as c0t3d0, be sure you label the disk so that VTOC gtes updated with the correct partition table. To recap, our scenario is: c0t3d0 (running Solaris 2.6) being copied to c0t1d0 (which will have the copied Solaris 2.6 slices/partitions) c0t3d0s0 / -> c0t1d0s0 / c0t3d0s4 /var -> c0t1d0s4 /var c0t3d0s5 /opt -> c0t1d0s5 /opt c0t3d0s6 /usr -> c0t1d0s6 /usr [6.] For each of the partitions that you wish to mount, run newfs to contruct a unix filesystem. So, newfs each partition. # newfs -v /dev/rdsk/c0t1d0s0 # newfs -v /dev/rdsk/c0t1d0s4 # newfs -v /dev/rdsk/c0t1d0s5 # newfs -v /dev/rdsk/c0t1d0s6 [7.] To ensure that they are clean and mounted properly, run fsck on these mounted partitions: # fsck /dev/rdsk/c0t1d0s0 # fsck /dev/rdsk/c0t1d0s4 # fsck /dev/rdsk/c0t1d0s5 # fsck /dev/rdsk/c0t1d0s6 [8.] Make the mount points. # mkdir /mount_point Create mountpoints for each slice/partition, like so: # mkdir /root2 # mkdir /var2 # mkdir /opt2 # mkdir /usr2 [9.] Mount the new partitions. # mount /dev/dsk/c0t1d0sX /mount_point Mount each partition (of the new disk), like so: # mount /dev/dsk/c0t1d0s0 /root2 # mount /dev/dsk/c0t1d0s4 /var2 # mount /dev/dsk/c0t1d0s5 /opt2 # mount /dev/dsk/c0t1d0s6 /usr2 [10.] Now we ufsdump each slices/partitions: It is often difficult to copy from one disk to another disk. If you try to use dd, and the disks are of differing sizes, then you will undoubtedly run into trouble. Use this method to copy from disk to disk and you should not have any problems. Of course you're still on the old disk (that's where you booted from c0t3d0): # cd / (Just ensures that you are in the root's parent/top directory). # ufsdump 0f - /dev/rdsk/c0t3d0s0 | (cd /root2; ufsrestore xf -) # ufsdump 0f - /dev/rdsk/c0t3d0s4 | (cd /var2; ufsrestore xf -) # ufsdump 0f - /dev/rdsk/c0t3d0s5 | (cd /opt2; ufsrestore xf -) # ufsdump 0f - /dev/rdsk/c0t3d0s6 | (cd /usr2; ufsrestore xf -) The gotcha here is that you can't really specify the directory name as ufsdump will interpret it as not being a block or character device. To illustrate this error: # cd /usr # ufsdump 0f - /usr | (cd /usr2; ufsrestore xf - ) DUMP: Writing 32 Kilobyte records DUMP: Date of this level 0 dump: Wed Dec 10 17:33:42 1997 DUMP: Date of last level 0 dump: the epoch DUMP: Dumping /dev/rdsk/c0t3d0s0 (tmpdns:/usr) to standard output. DUMP: Mapping (Pass I) [regular files] DUMP: Mapping (Pass II) [directories] DUMP: Estimated 317202 blocks (154.88MB). DUMP: Dumping (Pass III) [directories] DUMP: Broken pipe DUMP: The ENTIRE dump is aborted. However, using tar instead of ufsdump will work in this type of scenario: Example: # cd /usr # tar cvfp - . | (cd /usr2; tar xvfp - ) [11.] OPTIONAL (This may be redundant BUT ensures that the copied files are once again clean and consistent). Checking the integrity of a filesystem is always highly recommended even if it becomes redundant in nature. Now, check and run fsck on the new partition/slices: # fsck /dev/rdsk/c0t1d0s0 # fsck /dev/rdsk/c0t1d0s4 # fsck /dev/rdsk/c0t1d0s5 # fsck /dev/rdsk/c0t1d0s6 [12.] Edit your /mount_point/etc/vfstab file to have this disk bootup from the correct disk/devices c0t1d0 as opposed to c0t3d0. # cd /root2 # vi /root2/etc/vfstab Change c0tXd0sX devices to reflect the new disk! #device device mount FS fsck mount mount #to mount to fsck point type pass at boot options # #/dev/dsk/c1d0s2 /dev/rdsk/c1d0s2 /usr ufs 1 yes - fd - /dev/fd fd - no - /proc - /proc proc - no - /dev/dsk/c0t1d0s1 - - swap - no - /dev/dsk/c0t1d0s0 /dev/rdsk/c0t1d0s0 / ufs 1 no - /dev/dsk/c0t1d0s6 /dev/rdsk/c0t1d0s6 /usr ufs 1 no - /dev/dsk/c0t1d0s4 /dev/rdsk/c0t1d0s4 /var ufs 1 no - /dev/dsk/c0t1d0s5 /dev/rdsk/c0t1d0s5 /opt ufs 2 yes - swap - /tmp tmpfs - yes - :wq! [13.] Now you must run installboot to load a new bootblk on that disk. Not loading a bootblk will leave this disk in an unbootable state as the boot strap program is contained within the bootblk, and this in turn is what loads the boot file called ufsboot after interfacing with the OBP (Open Boot PROM). You can do this from your current booted disk or you may choose to boot off from cdrom via ok> boot cdrom -sw (single-user mode, writeable mode off of cdrom's mini-root). If you choose to get bootblk from your current disk: The location of the bootblk in Solaris 2.5 or higher is under: /usr/platform/`uname -i`/lib/fs/ufs/bootblk while in Solaris 2.4 or lower it's at: /usr/lib/fs/ufs/bootblk # /usr/sbin/installboot /usr/platform/`uname -i`/lib/fs/ufs/bootblk \ /dev/rdsk/c0t1d0s0 If you choose to get bootblk from your cdrom image: ok> boot cdrom -sw # installboot /cdrom/solaris_2_5_sparc/s0/export/exec/sparc.Solaris_2.5 \ /usr/platform/`uname -i`/lib/fs/ufs/bootblk /dev/rdsk/c0txd0s0 ANOTHER SPARC EXAMPLE: To install a ufs bootblock on slice 0 of target 0 on con- troller 1, of the platform where the command is being run, use: example# installboot /usr/platform/`uname -i`/lib/fs/ufs/bootblk \ /dev/rdsk/c1t0d0s0 [14.] Now create an alias for the other disk (this may be existent if it's off of the onboard/first scsi controller). ok> probe-scsi T3 original boot disk T1 new disk with copied slices Verify via devalias command to see current aliases: disk1 is for sd at 1,0 which is scsi id/target 1 ok> devalias ok> setenv boot-device disk1 ok> boot -rv (You do not necessarily need to do a reconfiguration boot as devices had already been created. This parameter will only be run if you attached new devices to your system). By default this will always boot from the new disk. If you want to boot from the old disk you can manually tell it to boot to that alias, like so: ok> boot disk or ok> boot disk3 (This will boot off from any Target 3/scsi id 3 internal disk). Also see INFODOC #'s 14046, 11855, 11854 for setting different boot devalias'es. NOTE: If the new disk encounters a problem on booting, most likely cause would be inappropriate devlinks so, the course of action to take here is the /etc/path_to_inst, /dev, /devices fix: The following is a solution to solve problems with /dev, /devices, and/or /etc/path-to_inst. This routine extracts the defaults (with links intact) from the Solaris 2.x CD-ROM. ok> boot cdrom -sw # mount /dev/dsk/c0t1d0s0 /a ** This step assumes your boot disk is c0t1d0s0 # cd /tmp/dev # tar cvfp - . | (cd /a/dev; tar xvfp - ) # cd /tmp/devices # tar cvfp - . | (cd /a/devices; tar xvfp - ) # cd /tmp/root/etc # cp path_to_inst /a/etc/path_to_inst # reboot -- -rv [15.] If you plan to move this new disk you copied the OS on, you MUST ensure that it will be moved to a similar architecture and machine type as hardware address paths are usually different from one machine to another. Each hardware platform has a hardware device tree which must match the device tree information saved during installation in /devices and the /dev directories. Another reason is that a kernel from one architecture cannot boot on a machine of a different architecture. Customers often overlook these architecture differences (Sun 4/4c/4m/4d/4u). A boot drive moved from a SPARCstation 2 (sun4c architecture) cannot boot on a SPARCstation 5 (sun4m architecture). For more details on why you can't move Solaris 2.X boot disk between machines please see INFODOC 13911 and 13920. Also ensure that you have the correct hostname, IP address and vfstab entries for this new drive if you plan to move it to another machine. APPLIES TO: Hardware, Operating Systems/Solaris/Solaris 2.x ATTACHMENTS: -----end of summary----- This afternoon, we spent about one hour switching the drives, and all went very well. Thanks to all, this list has been a great help! Peter. -- Peter Vos Easics NV - a TranSwitch Company Design Support Engineer System-on-Chip Design Tel: +32-16-395627 Interleuvenlaan 86 Fax: +32-16-395619 B-3001 Leuven - Belgium mailto:peterv at easics.be http://www.easics.com -------------- next part -------------- A non-text attachment was scrubbed... Name: smime.p7s Type: application/x-pkcs7-signature Size: 2915 bytes Desc: S/MIME Cryptographic Signature Url : http://www.sunmanagers.org/pipermail/summaries/attachments/20010913/ebdc3473/attachment.bin From Ken_Germann at bluecrossmn.com Thu Sep 13 13:28:45 2001 From: Ken_Germann at bluecrossmn.com (Ken_Germann at bluecrossmn.com) Date: Thu, 13 Sep 2001 12:28:45 -0500 Subject: Summary: applying Patches to a Jumpstart OS Image. Message-ID: Thanks to All fhose that replied to this inquiry: Ronald Loftin reloftin at syr.edu Dave Baker dave at dsb3.com Alexander Tomko Angel R. Rivera angel.r.rivera at usa.conoco.com Tim Evans tkevans at troweprice.com Frederic Medico I recieved 3 answers to the post: 1> Yes this can be done with a patchadd -C /export/jumpstart/OS/Solaris_7_1199-08/Solaris_2.7/Tools/Boot 2> #1 isn't recommended. The full OS is not being booted during a jumpstart.See Tim Evans reply below 3> use the WebFlash in Solaris 8 04/01. Tim Evans summarized about applying patches to the boot image: You misunderstand what a JumpStart image is. While it contains a minimal, bootable operating system that's used when client boot from it, it does not have "all" of Solaris installed as it would be on a running system. Solaris packages not needed to boot a jumpstart client are stored in pkgadd format, and get installed during a JumpStart. Some patches, therefore, will apply to a JumpStart image, but others will fail. In any case, though, *clients* won't get those patches installed just because you installed them on the boot server. They will temporarily boot a patched operating system, but not get those patches installed. You to do that after the fact, either manually, or with a custom finish script. _______________________________________________ sunmanagers mailing list sunmanagers at sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers From dkh at spod.net Thu Sep 13 17:16:35 2001 From: dkh at spod.net (Darren Honeyball) Date: Thu, 13 Sep 2001 22:16:35 +0100 Subject: SUMMARY: E10K SSP v3.4 Failover Message-ID: <00e801c13c99$5eb5f350$0132a8c0@defiant> Firstly, thanks to all who replied... This is the outcome of the problem... After several phone calls with Sun support, and still not getting anywhere, I finally found the problem, and it's a bug in one of Sun's scripts... ssp_config The nodename of the ssp was xxx-ssp1-uk-p, and the script uses the following line (line 2932 of ssp_config): typeset foo=$(egrep -l "^$(uname -n)$" /etc/hostname.*) Which basically does a grep of uname -n from /etc/hostname.*. uname -n in this case is xxx-ssp1-uk-p, and the hostname.* files contain xxx-ssp1-uk-p-eri0, xxx-ssp1-uk-p-qfe0 etc so the match actually finds multiple results, which make the next line: typeset adapt=${foo##*hostname.} fail. This strips of the hostname. part of the filename so you're left with an interface. So, I changed /etc/nodename from xxx-ssp1-uk-p to xxx-ssp1-uk-p-qfe2 and rebooted both ssp's. After a few seconds, the virtual interface is brought up automatically by the ssp software and everything is hunky dorey. (this made the nodename unique, and thus ssp_config would only find one hostname.* match) > Guys, > > I'm having trouble with the floating IP address in SSP 3.4 failover - has > anyone got this to work? > > It doesn't seem to start the virtual IP on the main ssp on boot, and also > doesn't move it across when you force a failover. > > Any ideas? Links to docs on setting this up? > > TIA, > > Darren > _______________________________________________ sunmanagers mailing list sunmanagers at sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagers From PZXKYS at atdva3.atd.gmeds.com Fri Sep 14 08:15:20 2001 From: PZXKYS at atdva3.atd.gmeds.com (Martin Meadows) Date: Fri, 14 Sep 2001 7:15:20 -0500 Subject: summary: directing unix time command output to log Message-ID: <010914071520.5b4e8@atdva3.atd.gmeds.com> Wow! I guess the question was too easy. I should have known this ... but had forgotten that time was going to stderr ... Lots of responses! Thanks to everyone for the excellent assistance! Marty Here's the original question: > I've been running a simple script test today & I wanted > to dump the output from this script to a log. I ran > the script and redirected the output to a log. The date > info (see below) is captured in the log. The result from > the "time" prefix (see below) on the mv command comes to > my screen instead of being written to the log. Can someone > explain how to get the time output to go to my log, too? > The script, in general, has something like this: > > date > time mv /ip4700/A0/scratch/testfile-eds4 /testfile-eds4 > date > time mv /testfile-eds4 /ip4700/A0/scratch/testfile-eds4 > date And here are a couple of the typical responses. -------------------------------- I think you probably need to redirect the output from stderr also. When you redirect using > only stdout is redirected. To redirect stderr you need something like this: script > log 2>&1 Where 'script' is the script you are running, 'log' is the log file. '2' is the handle for stderr and '1' is the handle for stdout so what '2>&1' does is redirect stderr to stdout. --------------------------------- Time sends its output to stderr; you need to redirect stderr. In csh: