cpu utilization question in FreeBSD
Hi, Im trying to put together a number of data recorders, to collect raw data from kernel or from applications. Here you can see a basic idea of SDR: http://www.systemdatarecorder.org/recording/recdesign.html http://www.systemdatarecorder.org/recording/recorders.html Im currently trying to port sysrec, http://www.systemdatarecorder.org/recording/sdr_bin/sysrec to FreeBSD. I would like to understand how CPU Utilization in FreeBSD counts, comparing to Solaris. Currently sysrec records the following metrics: # Utilisation, # CPU# usr + sys time across all CPUs # Memory # free RAM. freemem from availrmem # Disk # %busy. r+w times across all Disks # Network# throughput. r+w bytes across all NICs # # Saturation, # CPU# threads on the run queue # Memory # scan rate of the page scanner # Disk # operations on the wait queue # Network# errors due to buffer saturation # Would be fair to count CPU Util in FreeBSD similar with Solaris, User + System time ? Is there any Perl module or KSTAT functionality which exports these functionalities to userland for simple consumption ? Thanks, Stefan ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: Disabling openssl from ports
On 3 Feb 2010, at 03:36, Olivier Nicole wrote: > > I have one port, namely /usr/ports/www/pound that needs the version of > openssl from the ports (/usr/ports/security/openssl). > > But others ports works way better with the stock openssl from the > system. Personally, I've been using the ports version of openssl on a number of machines, and I haven't run into the sort of problems you claim. There is not a lot between the ports of the base system, especially if you're running a recent version of FreeBSD -- it's another port to manage, but you get access to various bits of new functionality. > Is there a configuration somewhere that could be used to say that > no-one except pound should use openssl from the ports? > > The only way I see is to put includes and libarries of openssl in some > obscure place and have pound point to them. > OK, this /should/ work. Add the following to /etc/make.conf: WITH_OPENSSL_BASE= yes .if ${.CURDIR:M*/www/pound} WITH_OPENSSL_PORT= yes .endif Test SSL-using executables with ldd(1) to see which copy of libcrypto they link against. Cheers, Matthew -- Dr Matthew J Seaman MA, D.Phil. 7 Priory Courtyard Flat 3 PGP: http://www.infracaninophile.co.uk/pgpkey Ramsgate Kent, CT11 9PW ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: can't make an 'a' slice except with auto-defaults
On Wednesday 03 February 2010 03:59:15 Steve Franks wrote: > On a running system. I mean, I know I should quit being a &%^#& and > read the manpage for bsdlabel, but sysintall really does have a nice > tui. sade(8) is the standalone version of sysinstall's partitioning subroutine. Also, if you're running a reasonably recent version of FreeBSD, you might want to take a look at gpart(8) which can do slicing and labeling (and a whole bunch of other disk partitioning related stuff). Regards, Pieter > 'C'reate slice goes straight to 'd', even on a 'fresh' disk. > I see in the handbook, this is alluded to, but some intermediate level > between begginer and expert (bsdlabel just strikes me as way too easy > to trash the disk I'm running off of while trying to make a backup), > would be nice...512M just won't fit the kernel+symbols. > > > > Steve ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Backup and FreeBSD/ZFS
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 I'm currently considering switching my Backend (Running WSS2008 Enterprise) to FreeBSD RELENG_8+zfs, however my last melee with ZFS and backups to Autoloader (HP 1/8 G2 LTO-3 job) didn't quite turn out in my favour. Seems FreeBSD has a decent backup solution for UFS/2 (dump/restore) but such luxuries where nowhere to be found for ZFS (but the performance was quite good!). So, my question is this: Can someone point me to the proper place to start reading on getting RELENG_8+zfs backed up to "tape robotics" (or more specifically: LTO-3 with a HP autoloader)? Will going the Opensolaris route be easier? (I had hoped to be able to use net/istgt/ from ports, but I guess I could find a different solution to that problem in Solaris). Any pointers would really help me here. The storage backend will be on MFI arrays (set up with redundant striping + automagic weekly consistency checks of the arrays). //Svein - -- - +---+--- /"\ |Svein Skogen | sv...@d80.iso100.no \ / |Solberg Østli 9| PGP Key: 0xE5E76831 X|2020 Skedsmokorset | sv...@jernhuset.no / \ |Norway | PGP Key: 0xCE96CE13 | | sv...@stillbilde.net ascii | | PGP Key: 0x58CD33B6 ribbon |System Admin | svein-listm...@stillbilde.net Campaign|stillbilde.net | PGP Key: 0x22D494A4 +---+--- |msn messenger: | Mobile Phone: +47 907 03 575 |sv...@jernhuset.no | RIPE handle:SS16503-RIPE - +---+--- If you really are in a hurry, mail me at svein-mob...@stillbilde.net This mailbox goes directly to my cellphone and is checked even when I'm not in front of my computer. - Picture Gallery: https://gallery.stillbilde.net/v/svein/ - -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.12 (MingW32) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAktpSlIACgkQODUnwSLUlKRvRQCcCKgCzTSCr9PVfyQ9cveGkuUd xTIAn2IWherBzlLTu/02CBLJMo34Ky2m =ruSv -END PGP SIGNATURE- ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: 8.0-RELEASE/amd64 - full ZFS install - low read and write disk performance
Dan Naumov wrote: > [j...@atombsd ~]$ dd if=/dev/zero of=/home/jago/test2 bs=1M count=4096 > 4096+0 records in > 4096+0 records out > 4294967296 bytes transferred in 143.878615 secs (29851325 bytes/sec) > > This works out to 1GB in 36,2 seconds / 28,2mb/s in the first test and > 4GB in 143.8 seconds / 28,4mb/s For the record, better results can be seen. In my test I put 3 Seagate Barracuda XT drives in a port multiplier and connected that to one port of a PCIe 3124 card. The MIRROR case is at about the I/O bandwidth limit of those drives. [r...@kraken ~]# zpool create tmpx ada{2,3,4} [r...@kraken ~]# dd if=/dev/zero of=/tmpx/test2 bs=1M count=4096 4096+0 records in 4096+0 records out 4294967296 bytes transferred in 20.892818 secs (205571470 bytes/sec) [r...@kraken ~]# zpool destroy tmpx [r...@kraken ~]# zpool create tmpx mirror ada{2,3} [r...@kraken ~]# dd if=/dev/zero of=/tmpx/test2 bs=1M count=4096 4096+0 records in 4096+0 records out 4294967296 bytes transferred in 36.432818 secs (117887321 bytes/sec) [r...@kraken ~]# ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: portsnap - broken metadata
On Sat, Jan 23, 2010 at 9:52 AM, Angelin Lalev wrote: > Greetings, > I believe I have broken the metainformation folders of portsnap. > How to fix them? > > ironholm# rm -rf /var/db/portsnap/ > ironholm# mkdir /var/db/portsnap > ironholm# portsnap fetch > Looking up portsnap.FreeBSD.org mirrors... 3 mirrors found. > Fetching public key from portsnap2.FreeBSD.org... done. > Fetching snapshot tag from portsnap2.FreeBSD.org... done. > Fetching snapshot metadata... done. > Fetching snapshot generated at Sat Jan 23 01:17:43 UTC 2010: > 8161b7b0a3b6b42453659f19197bfcc324b0a54b57dc29100% of 61 MB 249 kBps 00m00s > Extracting snapshot... done. > Verifying snapshot integrity... done. > Fetching snapshot tag from portsnap2.FreeBSD.org... done. > Fetching snapshot metadata... done. > Updating from Sat Jan 23 01:17:43 UTC 2010 to Sat Jan 23 08:00:35 UTC 2010. > Fetching 4 metadata patches. done. > Applying metadata patches... done. > Fetching 4 metadata files... /usr/sbin/portsnap: cannot open > bd5906dc86367765516942be65b56170d979598ac1325709aa83e67efec39d6d.gz: > No such file or directory > metadata is corrupt. > ironholm# > I have some additional details: I have proxy server that requires username and password and I have set the HTTP_PROXY and FTP_PROXY environment variables accordingly. (HTTP_PROXY = FTP_PROXY = USERNAME:passw...@proxy.uni-svishtov.bg:8080) portsnap --debug shows, among other rows the following error message: phttpget: host = USERNAME, port = passw...@proxy.uni-svishtov.bg:8080: servname not supported for ai_socktype Seems to me like a bug somewhere in portsnap. Any help? ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: can't make an 'a' slice except with auto-defaults
Just a sidenote to avoid misunderstandings: On Tue, 2 Feb 2010 19:59:15 -0700, Steve Franks wrote: > 'C'reate slice goes straight to 'd', even on a 'fresh' disk. I think you create a partition, not a slice. The slice editor would 'C'reate 's1'. :-) > I see in the handbook, this is alluded to, but some intermediate level > between begginer and expert (bsdlabel just strikes me as way too easy > to trash the disk I'm running off of while trying to make a backup), > would be nice...512M just won't fit the kernel+symbols. According to the handbook, http://www.freebsd.org/doc/handbook/install-steps.html you should make sure that you've selected a CUSTOM install, visit the slice editor (if needed) and the partition editor afterwards. Creating partitions should start with 'a', and if you want to use the whole disk, 'a' is the default (because 'a' usually refers to the booting partition). A case when no 'a' partition can be created does exist when there is already an 'a' partition on the slice. If that's the case, make sure it's not the case. :-) > FUMEN PROHIB, LEGERE LEVITICUS! :-) -- Polytropon Magdeburg, Germany Happy FreeBSD user since 4.0 Andra moi ennepe, Mousa, ... ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: Disabling openssl from ports
>OK, this /should/ work. Add the following to /etc/make.conf: > >WITH_OPENSSL_BASE= yes > >.if ${.CURDIR:M*/www/pound} >WITH_OPENSSL_PORT= yes >.endif No, it won't -- at least, if you leave it in make.conf after building www/pound, it wil break all subsequent rebuilds of all other ports that depend upon the base system openssl. Matthew, you ought to know better ... >From bsd.openssl.mk: .if defined(WITH_OPENSSL_BASE) ... .if exists(${LOCALBASE}/lib/libcrypto.so) check-depends:: @${ECHO_CMD} "Dependency error: this port wants the OpenSSL library from the FreeBSD" @${ECHO_CMD} "base system. You can't build against it, while a newer" @${ECHO_CMD} "version is installed by a port." @${ECHO_CMD} "Please deinstall the port or undefine WITH_OPENSSL_BASE." @${FALSE} .endif Mixing and matching the different openssl versions can lead to problems (for one thing, there are too many sloppy LDFLAGS=-L${LOCALBASE}/lib floating around in different ports), and you'll have to hack port Makefiles and use ldd(1) or other tools to verify that your changes work. You're probably better off just using one or the other. If you still want to try it, then I suggest installing security/openssl in non-default PREFIX, then patching the www/pound Makefile so that it doesn't use USE_OPENSSL, and then adding whatever variables are needed by it's configure script to locate and link with security/openssl to CONFIGURE_ENV and/or MAKE_ENV, as well as the proper LIB_DEPENDS on security/openssl. After doing this and installing www/pound, if rtld(1) is still loading the base system openssl when www/pound binaries are executed, or can't find the security/openssl libraries off in their non-default location, then use libmap.conf(5) to point (only) the www/pound binaries to the security/openssl libraries. You'll have to ensure that your changes to www/pound's Makefile aren't wiped out by subsequent updates to your Ports tree, of course. b. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Helps! Installation hangs for FreeBSD 4.x and 5.1
Dear All, I have encountered in installing FreeBSD including version 4.3, 4.5, 4.6, 4.8 and 5.1 on my new Dell laptop. Here is some info about the laptop. It came with Vista Home Basic. I have formatted the hard drive and installed XP on it. While doing so, I change the SATA option from AHCI to ATA, which may cause my following problem. When I tried to install various version of FreeBSD, installation hangs at different points. For 4.3 and 4.5, installation hangs at following line: *ppc0: parallel port not found* For 4.6 and 4.8, installation hangs at following line: *ata1-slave: ATAPI identify retries exceeded* For 5.1, installation printed out something like: *:/ write failed, filesystem is full panic: Going nowhere without my init! syncing disks, buffer remaining ... 454 454 290 290 290 giving up on 125 buffers* I have never encountered such problem since I have been playing with 4.3 for a while on other older computers. I am wondering if any one has similar experience and how it could be sovled. Thank all of you for your time to read my question. Millions of thanks to all you guys! Your sincerely, Paul Shi Electronic and Communication Engineering Senior Department of Electrical and Electronic Engineering University of Hong Kong ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: portsnap - broken metadata
On Wed, 3 Feb 2010 12:20:53 + Angelin Lalev wrote: > I have proxy server that requires username and password and I have set > the HTTP_PROXY and FTP_PROXY environment variables accordingly. > (HTTP_PROXY = FTP_PROXY = > USERNAME:passw...@proxy.uni-svishtov.bg:8080) > > portsnap --debug shows, among other rows the following error message: > > phttpget: host = USERNAME, port = passw...@proxy.uni-svishtov.bg:8080: > servname not > supported for ai_socktype > > Seems to me like a bug somewhere in portsnap. Any help? It looks to be a bug in phttpget, which is a download utility used by portsnap and freebsd-update. It supports pipelining, so it's very fast at fetching a large number of small files, but in all other respects it simpler and less mature than fetch. As temporary workaround I would suggest you backup portsnap and change the line: PHTTPGET=/usr/libexec/phttpget to PHTTPGET=/usr/bin/fetch ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: Helps! Installation hangs for FreeBSD 4.x and 5.1
On 02/03/10 14:22, Paul Shi wrote: Dear All, I have encountered in installing FreeBSD including version 4.3, 4.5, 4.6, 4.8 and 5.1 on my new Dell laptop. Here is some info about the laptop. These versions of FreeBSD are so old they probably don't support your hardware. Try installing FreeBSD 8.0. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: portsnap - broken metadata
On Wed, 3 Feb 2010 13:22:59 + RW wrote: > On Wed, 3 Feb 2010 12:20:53 + > Angelin Lalev wrote: > > > > I have proxy server that requires username and password and I have > > set the HTTP_PROXY and FTP_PROXY environment variables accordingly. > > (HTTP_PROXY = FTP_PROXY = > > USERNAME:passw...@proxy.uni-svishtov.bg:8080) > > > > portsnap --debug shows, among other rows the following error > > message: > > > > phttpget: host = USERNAME, port = > > passw...@proxy.uni-svishtov.bg:8080: servname not > > supported for ai_socktype > > > > Seems to me like a bug somewhere in portsnap. Any help? > > It looks to be a bug in phttpget, which is a download utility > used by portsnap and freebsd-update. It supports pipelining, so > it's very fast at fetching a large number of small files, but in all > other respects it simpler and less mature than fetch. > > As temporary workaround I would suggest you backup portsnap and change > the line: > > PHTTPGET=/usr/libexec/phttpget > > to > > PHTTPGET=/usr/bin/fetch I see that isn't going to work since phttpget needs it's arguments in the form phttpget server file1 file2 ... fileN you'd need to wrap fetch if you want to try that. Anyway, I've CC'd Colin Percival which I forgot to do before, and I suggest you open a PR. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
GeForce GTX 260M on amd64
Hello, I have bought new notebook with GeForce GTX 260M and installed amd64 distribution of freebsd 8 because of 4GB RAM. Is there any way to run OpenGL (get drivers for this graphics card)? Thanks Radek ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: fixing up port dependencies properly
John W wrote: >I updated my ports tree with csup, and tried to run 'portmaster -na'. >It gave me this: > >===>>> The mail/p5-Email-Simple-Creator port has been deleted: >Folded into p5-Email-Simple package > >Ok, that makes sense. But what do I do to fix it? >It seems I need to replace dependencies on p5-Email-Simple-Creator >with dependencies on p5-Email-Simple. > >But if I manually do that, won't my changes be blown away the next >time I update ports? The committer who added the entry to /usr/ports/MOVED also seems to have adjusted any dependencies in the Ports tree, back on 24 Nov. 2009. So if you have an up-to-date ports tree, then after rebuilding the ports that used to depend upon p5-Email-Simple-Creator, those ports will depend instead upon p5-Email-Simple, and no further intervention will be needed. > >Perhaps I should use the '-o' (origin) option of portmaster? I'm not >100% sure what that does, incidentally (explanation welcome). >I assume something like: > >portmaster -o p5-Email-Simple p5-Email-Simple-Creator > I don't use portmaster often, but I think it should instead be: portmaster -o mail/p5-Email-Simple p5-Email-Simple-Creator Read the portmaster(1) manpage carefully, and look at the examples. >Will those changes get blown away by the next update of ports? In this case, no. > >Is the most correct solution just to wait until all maintainers of >ports which depend on p5-Email-Simple-Creator each update their >makefiles to depend on p5-Email-Simple, instead? (Though that doesn't >help in the short term :) That should already have been done. In other cases, if it has not, then you should send a message to the committer who made the change (if you aren't familiar with cvs(1), which is used to manage the ports repository, then you can use cvsweb.freebsd.org or www.freshports.org to find this information), and to the maintainers of the ports that have the outdated dependencies. If they don't respond within a reasonable amount of time, then file a Problem Report: http://www.freebsd.org/support/bugreports.html While they are fixing the problems, you can patch the dependent ports yourself (this is sometimes as simple as changing the *_DEPENDS line in the port Makefile; other times, it requires patches to the port sources), and then rebuild the ports; or you can try to use portmaster -o , or portupgrade -o, which will succeed in the simplest cases. You could also do it manually, by using sed(1) to substitute every occurrence of the old PKGNAME with the new PKGNAME in the @pkgdep lines in /var/db/pkg/*/+COMMENTS, and likewise for the PKGORIGIN values preceded by DEPORIGIN. However, be careful when tinkering with /var/db/pkg -- you should back it up first before making changes. b. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: fixing up port dependencies properly
On 2/3/10, b. f. wrote: > John W wrote: > -o , or portupgrade -o, which will succeed in the simplest cases. You > could also do it manually, by using sed(1) to substitute every > occurrence of the old PKGNAME with the new PKGNAME in the @pkgdep > lines in /var/db/pkg/*/+COMMENTS, and likewise for the PKGORIGIN Sorry, that should be /var/db/pkg/*/+CONTENTS, of course. Where is my first cup of coffee.. b. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: Disabling openssl from ports
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 03/02/2010 12:57, b. f. wrote: >> OK, this /should/ work. Add the following to /etc/make.conf: >> > >> >WITH_OPENSSL_BASE= yes >> > >> >.if ${.CURDIR:M*/www/pound} >> >WITH_OPENSSL_PORT= yes >> >.endif > No, it won't -- at least, if you leave it in make.conf after building > www/pound, it wil break all subsequent rebuilds of all other ports > that depend upon the base system openssl. Matthew, you ought to know > better ... That's what I get for not testing. In fact, it doesn't work at all -- pound gets linked against the base system openssl. That's because 'WITH_OPENSSL_BASE' is defined, and that takes precedence over 'WITH_OPENSSL_PORT'. If I fix that, then, yes, you can't install any ports subsequently that link against the base OpenSSL. Cheers, Matthew - -- Dr Matthew J Seaman MA, D.Phil. 7 Priory Courtyard Flat 3 PGP: http://www.infracaninophile.co.uk/pgpkey Ramsgate Kent, CT11 9PW -BEGIN PGP SIGNATURE- Version: GnuPG/MacGPG2 v2.0.14 (Darwin) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAktpiL8ACgkQ8Mjk52CukIw1awCgg4zuEIwkgyHunvykfmAwAofr phwAninofTbkasi39SoHMIlu7Hr2M4qc =qz7n -END PGP SIGNATURE- ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: GeForce GTX 260M on amd64
Hi Did you try the binary driver x11/nvidia-driver? Your card should be supported. 2010/2/3 Bc. Radek Krejca : > Hello, > > I have bought new notebook with GeForce GTX 260M and installed amd64 > distribution of freebsd 8 because of 4GB RAM. > > Is there any way to run OpenGL (get drivers for this graphics card)? > > Thanks > Radek > > ___ > freebsd-questions@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-questions > To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org" > ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: can't make an 'a' slice except with auto-defaults
On Tue, Feb 02, 2010 at 07:59:15PM -0700, Steve Franks wrote: > On a running system. I mean, I know I should quit being a &%^#& and > read the manpage for bsdlabel, but sysintall really does have a nice > tui.'C'reate slice goes straight to 'd', even on a 'fresh' disk. > I see in the handbook, this is alluded to, but some intermediate level > between begginer and expert (bsdlabel just strikes me as way too easy > to trash the disk I'm running off of while trying to make a backup), > would be nice...512M just won't fit the kernel+symbols. > > Well, Create slice would be an fdisk(8) thing, not bsdlabel. bsdlabel creates partitions within a slice. But, generally you cannot run fdisk on a disk that is in use on a running system - which generally means that it is the boot device, has filesystems mounted or has part of the currently designated swap space. You will need to plug in a boot cd or bring up the fixit system for that. The fixit system runs from memory - creates filesystems and mount points in memory rather than on disk, so it can talk to any disk. New, if you are working on a non-used (extra) disk, eg one that is not the boot device nor has any mounted filesystems or swap space on it, then you should be able to fdisk and bsdlabel that from a running system. I have no idea what you mean by "'C'reate slice goes straight to 'd'" It does not match anything I remember being possible. I don't happen to have any system handy at the moment that I can muck with disks on. jerry > > Steve > ___ > freebsd-questions@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-questions > To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org" ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: Backup and FreeBSD/ZFS
On Wed, Feb 03, 2010 at 11:05:06AM +0100, Svein Skogen (Listmail Account) wrote: > -BEGIN PGP SIGNED MESSAGE- > Hash: SHA1 > > I'm currently considering switching my Backend (Running WSS2008 > Enterprise) to FreeBSD RELENG_8+zfs, however my last melee with ZFS and > backups to Autoloader (HP 1/8 G2 LTO-3 job) didn't quite turn out in my > favour. > > Seems FreeBSD has a decent backup solution for UFS/2 (dump/restore) but > such luxuries where nowhere to be found for ZFS (but the performance was > quite good!). I haven't used ZFS yet, but if stuff is put there in 'file' with an inode for each, I wonder if dump/restore would actually work on ZFS. Of course, it would not preserve the formatting/filesystem building. But it doesn't do that fur UFS either. Backups created by dump and read by restore are just files with a series of files as they come from disk organized and located by inode. But, as I say, I haven't worked with ZFS yet so do not know how date is kept track of on ZFS. jerry > > So, my question is this: Can someone point me to the proper place to > start reading on getting RELENG_8+zfs backed up to "tape robotics" (or > more specifically: LTO-3 with a HP autoloader)? Will going the > Opensolaris route be easier? (I had hoped to be able to use net/istgt/ > from ports, but I guess I could find a different solution to that > problem in Solaris). > > Any pointers would really help me here. The storage backend will be on > MFI arrays (set up with redundant striping + automagic weekly > consistency checks of the arrays). > > //Svein > > - -- > - +---+--- > /"\ |Svein Skogen | sv...@d80.iso100.no > \ / |Solberg Østli 9| PGP Key: 0xE5E76831 >X|2020 Skedsmokorset | sv...@jernhuset.no > / \ |Norway | PGP Key: 0xCE96CE13 > | | sv...@stillbilde.net > ascii | | PGP Key: 0x58CD33B6 > ribbon |System Admin | svein-listm...@stillbilde.net > Campaign|stillbilde.net | PGP Key: 0x22D494A4 > +---+--- > |msn messenger: | Mobile Phone: +47 907 03 575 > |sv...@jernhuset.no | RIPE handle:SS16503-RIPE > - +---+--- > If you really are in a hurry, mail me at >svein-mob...@stillbilde.net > This mailbox goes directly to my cellphone and is checked > even when I'm not in front of my computer. > - > Picture Gallery: > https://gallery.stillbilde.net/v/svein/ > - > -BEGIN PGP SIGNATURE- > Version: GnuPG v2.0.12 (MingW32) > Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ > > iEYEARECAAYFAktpSlIACgkQODUnwSLUlKRvRQCcCKgCzTSCr9PVfyQ9cveGkuUd > xTIAn2IWherBzlLTu/02CBLJMo34Ky2m > =ruSv > -END PGP SIGNATURE- > ___ > freebsd-questions@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-questions > To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org" > ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: crontab
Nadir Aliyev writes: > Hello friends. > > I have interesting situation with cron. > > I created a > simple script for process monitoring: > > #!/usr/local/bin/bash > processname=`/bin/ps aux | /usr/bin/grep -v grep | > /usr/bin/grep -c 'maintenance_jobs.php'` > if [ $processname -le "0" ]; > then > echo "`/bin/date` > JOB WAS DEAD. RESTARTED!" | mail -s "ATTENTION" > m...@email.net; > /usr/local/bin/php > /usr/local/www/web/bin/maintenance_jobs.php then" is not interpretated by > shell when i run this script from cron. > > I tried it on sh and bash. Result > is same. > But this script worked on pre 8 versions. The script got wrapped and apparently cut off. I can't understand it, and probably nobody else could either. -- Lowell Gilbert, embedded/networking software engineer, Boston area http://be-well.ilk.org/~lowell/ ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Mac applications on FreeBSD
Will Mac OS X applications run on FreeBSD? Specifically, Microsoft Office:Mac? If so, does it require special manipulation like hacking the system kernel and/or application in order to get it to run? Thanks very much, Bill ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: Mac applications on FreeBSD
Hi, On 03 February 2010 pm 23:38:55 Bill White wrote: > Will Mac OS X applications run on FreeBSD? Specifically, Microsoft > Office:Mac? If so, does it require special manipulation like hacking > the system kernel and/or application in order to get it to run? Thanks not to my knowledge. Only some Windows applications can be used on FreeBSD via wine. Erich ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: Mac applications on FreeBSD
On 03/02/2010 15:38, Bill White wrote: > Will Mac OS X applications run on FreeBSD? Specifically, Microsoft > Office:Mac? If so, does it require special manipulation like hacking > the system kernel and/or application in order to get it to run? Thanks > very much, Bill No. FreeBSD can run many Linux applications under emulation (and possibly some Sco applications, but that code branch has rotted due to disuse). It can't run Mac OS X applications though. Despite certain similarities between FreeBSD and the unix underpinnings of Mac OS X (darwin) i's not possible to run even a command-line program from Mac OS X on a FreeBSD box. Not that that's a particularly interesting thing to do. Obviously what you want to run are all those nice graphical apps written in cocoa and whaterver else is flavour of the month at Apple. There is simply *no* chance of that: the graphical environment is Apples' proprietary code, and they aren't going to be releasing it any time soon. Without the graphical layers, trying to get a Mac application running is an exercise in futility. Why don't you explain what exactly you're trying to achieve: free alternatives to most Mac applications are generally available. They won't be anything like as slick, and functionality will only be approximately the same, but they might do. Cheers, Matthew (who is getting fed up with Thunderbird3 continually marking whole mailboexs as unread on his *shiny* new Mac...) -- Dr Matthew J Seaman MA, D.Phil. 7 Priory Courtyard, Flat 3 Black Earth Consulting Ramsgate Kent, CT11 9PW Free and Open Source Solutions Tel: +44 (0)1843 580647 signature.asc Description: OpenPGP digital signature
Re: Backup and FreeBSD/ZFS
I have been using Amanda straight from the port, it supports both tar from snapshots if you need to be able to retrieve individual files from the backup and zfs send if recovery at filesystem level is OK. --glz --On Wednesday, February 03, 2010 11:05 AM +0100 "Svein Skogen (Listmail Account)" wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 I'm currently considering switching my Backend (Running WSS2008 Enterprise) to FreeBSD RELENG_8+zfs, however my last melee with ZFS and backups to Autoloader (HP 1/8 G2 LTO-3 job) didn't quite turn out in my favour. Seems FreeBSD has a decent backup solution for UFS/2 (dump/restore) but such luxuries where nowhere to be found for ZFS (but the performance was quite good!). So, my question is this: Can someone point me to the proper place to start reading on getting RELENG_8+zfs backed up to "tape robotics" (or more specifically: LTO-3 with a HP autoloader)? Will going the Opensolaris route be easier? (I had hoped to be able to use net/istgt/ from ports, but I guess I could find a different solution to that problem in Solaris). Any pointers would really help me here. The storage backend will be on MFI arrays (set up with redundant striping + automagic weekly consistency checks of the arrays). //Svein - -- - +---+--- /"\ |Svein Skogen | sv...@d80.iso100.no \ / |Solberg Østli 9| PGP Key: 0xE5E76831 X|2020 Skedsmokorset | sv...@jernhuset.no / \ |Norway | PGP Key: 0xCE96CE13 | | sv...@stillbilde.net ascii | | PGP Key: 0x58CD33B6 ribbon |System Admin | svein-listm...@stillbilde.net Campaign|stillbilde.net | PGP Key: 0x22D494A4 +---+--- |msn messenger: | Mobile Phone: +47 907 03 575 |sv...@jernhuset.no | RIPE handle:SS16503-RIPE - +---+--- If you really are in a hurry, mail me at svein-mob...@stillbilde.net This mailbox goes directly to my cellphone and is checked even when I'm not in front of my computer. - Picture Gallery: https://gallery.stillbilde.net/v/svein/ - -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.12 (MingW32) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAktpSlIACgkQODUnwSLUlKRvRQCcCKgCzTSCr9PVfyQ9cveGkuUd xTIAn2IWherBzlLTu/02CBLJMo34Ky2m =ruSv -END PGP SIGNATURE- ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org" ... the future isMobile Goran Lowkrantz System Architect, isMobile, Aurorum 2, S-977 75 Luleå, Sweden Phone: +46(0)920-75559 Mobile: +46(0)70-587 87 82 Fax: +46(0)70-615 87 82 http://www.ismobile.com ... ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: can't make an 'a' slice except with auto-defaults
On Wed, Feb 3, 2010 at 8:18 AM, Jerry McAllister wrote: > On Tue, Feb 02, 2010 at 07:59:15PM -0700, Steve Franks wrote: > >> On a running system. I mean, I know I should quit being a &%^#& and >> read the manpage for bsdlabel, but sysintall really does have a nice >> tui. 'C'reate slice goes straight to 'd', even on a 'fresh' disk. >> I see in the handbook, this is alluded to, but some intermediate level >> between begginer and expert (bsdlabel just strikes me as way too easy >> to trash the disk I'm running off of while trying to make a backup), >> would be nice...512M just won't fit the kernel+symbols. >> >> > > Well, Create slice would be an fdisk(8) thing, not bsdlabel. > bsdlabel creates partitions within a slice. > > But, generally you cannot run fdisk on a disk that is in use on a > running system - which generally means that it is the boot device, > has filesystems mounted or has part of the currently designated swap > space. You will need to plug in a boot cd or bring up the fixit system > for that. The fixit system runs from memory - creates filesystems > and mount points in memory rather than on disk, so it can talk to > any disk. > > New, if you are working on a non-used (extra) disk, eg one that is not > the boot device nor has any mounted filesystems or swap space > on it, then you should be able to fdisk and bsdlabel that from > a running system. > > I have no idea what you mean by "'C'reate slice goes straight to 'd'" > It does not match anything I remember being possible. I don't happen > to have any system handy at the moment that I can muck with disks on. > > jerry Ok, terminology crash. As someone pointed out, I'm talking about label, here, not fdisk, and partitions, not slices (had those two backwards in my head). Basically, as far as I can tell, on a running system, there is no combination of keystrokes in sysinstall's label editor that will create an "ad[1-9]s1a", except the 'a' key which produces a 512M s1a. All other keystrokes (namely 'c') go straight to "ad[1-9]s1d" when a second disk is placed in a system booted from ad0s1a. I'm just trying to make a fresh disk ready for dump/restore with a 1G /, so I guess sysinstall is out as an option at this time. Steve ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
adding disk moves ad0 to ad4
Just curious, having read the handbook section talking about freebsd going straight to the hardware and skipping the bios for disk numbering, why then, if I stick a sata disk in 'sata0' on the motherboard, does it come up as ad0, but if I add a second disk in 'sata1' or 'pata0', on the next boot, I have no ad0, but ad4 and ad6? This seems to be the case with every mobo I've owned in the last 2 years from a variety of mfr's. Is there a way around this? I don't care what it comes out as, as long as it stays put... Since I have the only fbsd system at work, I tend to format alot of funky drives for people, and it gets anoying having to swap fstab's every time... Steve ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: can't make an 'a' slice except with auto-defaults
On Wed, 3 Feb 2010 10:26:06 -0700, Steve Franks wrote: > I'm just trying > to make a fresh disk ready for dump/restore with a 1G /, so I guess > sysinstall is out as an option at this time. Why so complicated? The command # newfs /dev/ad1 will prepare the disk, assuming ad1 is the new disk. It will create one single partition covering the whole disk, with no slice. You can then run # mount /dev/ad1 /mnt to access it; /dev/ad1 is the same as /dev/ad1c, "the whole disk". By the way, have you tried the program # sade instead of sysinstall? -- Polytropon Magdeburg, Germany Happy FreeBSD user since 4.0 Andra moi ennepe, Mousa, ... ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: adding disk moves ad0 to ad4
On Wed, Feb 3, 2010 at 11:35 AM, Steve Franks wrote: > Just curious, having read the handbook section talking about freebsd > going straight to the hardware and skipping the bios for disk > numbering, why then, if I stick a sata disk in 'sata0' on the > motherboard, does it come up as ad0, but if I add a second disk in > 'sata1' or 'pata0', on the next boot, I have no ad0, but ad4 and ad6? > This seems to be the case with every mobo I've owned in the last 2 > years from a variety of mfr's. Is there a way around this? I don't > care what it comes out as, as long as it stays put... Since I have > the only fbsd system at work, I tend to format alot of funky drives > for people, and it gets anoying having to swap fstab's every time... > > Steve > Use glabel, then you can used the same fstab across disk swaps/machinces -- Adam Vande More ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: can't make an 'a' slice except with auto-defaults
> Why so complicated? The command > > # newfs /dev/ad1 I'm looking to mirror/dup/image the entire system to something I can stick in another system.I hear there's good reasons for not running my whole system off of a single partition. The 'other' system has 7.2 and has devolved to a 25% chance of a hard freeze every time I unplug a ucom device (seems to have cropped up between 7.2-release and 7.2-stable#3). 8 likes usb, so I like 8. Steve ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: can't make an 'a' slice except with auto-defaults
On Wed, 3 Feb 2010 10:42:42 -0700, Steve Franks wrote: > > Why so complicated? The command > > > > # newfs /dev/ad1 > > I'm looking to mirror/dup/image the entire system to something I can > stick in another system. Sorry, my misunderstanding. I thought you're intending to use the second disk as a pure data disk. > I hear there's good reasons for not > running my whole system off of a single partition. The example above would not only create one single partition, it would furthermore omit the slice containing it. :-) -- Polytropon Magdeburg, Germany Happy FreeBSD user since 4.0 Andra moi ennepe, Mousa, ... ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: adding disk moves ad0 to ad4
On Wed, 3 Feb 2010 10:35:48 -0700, Steve Franks wrote: > Just curious, having read the handbook section talking about freebsd > going straight to the hardware and skipping the bios for disk > numbering, why then, if I stick a sata disk in 'sata0' on the > motherboard, does it come up as ad0, but if I add a second disk in > 'sata1' or 'pata0', on the next boot, I have no ad0, but ad4 and ad6? > This seems to be the case with every mobo I've owned in the last 2 > years from a variety of mfr's. Is there a way around this? Maybe this is specific to your motherboard. As far as I experienced, using (P)ATA and SATA - or not using it - keeps the numbering intact, e. g. ad0 - ad3 is ATA, ad4 - ad7 is SATA, no matter where a disk is actually connected. It's possible that your BIOS does something strange in representing one SATA, but no ATA disk as ad0, "the first disk existing", as well as if an ATA disk would be present, but no SATA disk. I can understand that this is annoying. > I don't > care what it comes out as, as long as it stays put... There are labels or UFSIDs you can use to identify partitions on a disk regardless of the device name they would come out as. See "man glabel" and "man tunefs" for details. > Since I have > the only fbsd system at work, I tend to format alot of funky drives > for people, and it gets anoying having to swap fstab's every time... Labels can really help here: /etc/fstab will then contain labels or UFSIDs instead of device names - and they don't change when a disk is added or removed. -- Polytropon Magdeburg, Germany Happy FreeBSD user since 4.0 Andra moi ennepe, Mousa, ... ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: adding disk moves ad0 to ad4
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 03/02/2010 17:35, Steve Franks wrote: > Just curious, having read the handbook section talking about freebsd > going straight to the hardware and skipping the bios for disk > numbering, why then, if I stick a sata disk in 'sata0' on the > motherboard, does it come up as ad0, but if I add a second disk in > 'sata1' or 'pata0', on the next boot, I have no ad0, but ad4 and ad6? > This seems to be the case with every mobo I've owned in the last 2 > years from a variety of mfr's. Is there a way around this? I don't > care what it comes out as, as long as it stays put... Since I have > the only fbsd system at work, I tend to format alot of funky drives > for people, and it gets anoying having to swap fstab's every time... I've never seen disks spontaneously changing from ad0 to ad4. I have seen motherboards where toggling BIOS settings to do with PATA compatibility / SATA support changes the device number. Sata drives frequently appear as ad4 and ad6 -- that's because each SATA connector is behaving like a separate IDE bus (with 2 devices -- master and slave -- allocated per bus, even though you'ld only ever use one with SATA disks). Devices ad0 -- ad3 are allocated to the *real* IDE busses on the motherboard -- probably one of those slots is taken up by a CD or DVD drive. Now, with some mobo's ad4 and ad6 is all you're going to get. Others, you can tweak some settings in the BIOS and your SATA drives will then appear as ad0 and ad1. Unfortunately it's pot luck as to what kit this works with, and what it doesn't. Cheers, Matthew - -- Dr Matthew J Seaman MA, D.Phil. 7 Priory Courtyard Flat 3 PGP: http://www.infracaninophile.co.uk/pgpkey Ramsgate Kent, CT11 9PW -BEGIN PGP SIGNATURE- Version: GnuPG/MacGPG2 v2.0.14 (Darwin) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAktpuIwACgkQ8Mjk52CukIy8MACfdua7RHPUT+qTKtc608W2pPWk /0YAniWJVjPogJEolV+maVHit7xTpYXm =WB7h -END PGP SIGNATURE- ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: Backup and FreeBSD/ZFS
If my memory serves, you're looking at something similar to taking a snapshot and then sending it to the tape device, so zfs send | (tape device access). This, IIRC, is functionally identical to dump/restore. The Solaris ZFS Admin guide is generally helpful (even as we live in the FreeBSD world). (http://docs.sun.com/app/docs/doc/819-5461/6n7ht6qsc?a=view) On 2/3/2010 4:05 AM, Svein Skogen (Listmail Account) wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 I'm currently considering switching my Backend (Running WSS2008 Enterprise) to FreeBSD RELENG_8+zfs, however my last melee with ZFS and backups to Autoloader (HP 1/8 G2 LTO-3 job) didn't quite turn out in my favour. Seems FreeBSD has a decent backup solution for UFS/2 (dump/restore) but such luxuries where nowhere to be found for ZFS (but the performance was quite good!). So, my question is this: Can someone point me to the proper place to start reading on getting RELENG_8+zfs backed up to "tape robotics" (or more specifically: LTO-3 with a HP autoloader)? Will going the Opensolaris route be easier? (I had hoped to be able to use net/istgt/ from ports, but I guess I could find a different solution to that problem in Solaris). Any pointers would really help me here. The storage backend will be on MFI arrays (set up with redundant striping + automagic weekly consistency checks of the arrays). //Svein - -- - +---+--- /"\ |Svein Skogen | sv...@d80.iso100.no \ / |Solberg Østli 9| PGP Key: 0xE5E76831 X|2020 Skedsmokorset | sv...@jernhuset.no / \ |Norway | PGP Key: 0xCE96CE13 | | sv...@stillbilde.net ascii | | PGP Key: 0x58CD33B6 ribbon |System Admin | svein-listm...@stillbilde.net Campaign|stillbilde.net | PGP Key: 0x22D494A4 +---+--- |msn messenger: | Mobile Phone: +47 907 03 575 |sv...@jernhuset.no | RIPE handle:SS16503-RIPE - +---+--- If you really are in a hurry, mail me at svein-mob...@stillbilde.net This mailbox goes directly to my cellphone and is checked even when I'm not in front of my computer. - Picture Gallery: https://gallery.stillbilde.net/v/svein/ - -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.12 (MingW32) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAktpSlIACgkQODUnwSLUlKRvRQCcCKgCzTSCr9PVfyQ9cveGkuUd xTIAn2IWherBzlLTu/02CBLJMo34Ky2m =ruSv -END PGP SIGNATURE- ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org" ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: adding disk moves ad0 to ad4
On Wed, Feb 3, 2010 at 10:54 AM, Polytropon wrote: > On Wed, 3 Feb 2010 10:35:48 -0700, Steve Franks > wrote: >> Just curious, having read the handbook section talking about freebsd >> going straight to the hardware and skipping the bios for disk >> numbering, why then, if I stick a sata disk in 'sata0' on the >> motherboard, does it come up as ad0, but if I add a second disk in >> 'sata1' or 'pata0', on the next boot, I have no ad0, but ad4 and ad6? >> This seems to be the case with every mobo I've owned in the last 2 >> years from a variety of mfr's. Is there a way around this? > > Maybe this is specific to your motherboard. As far as I > experienced, using (P)ATA and SATA - or not using it - > keeps the numbering intact, e. g. ad0 - ad3 is ATA, > ad4 - ad7 is SATA, no matter where a disk is actually > connected. > > It's possible that your BIOS does something strange in > representing one SATA, but no ATA disk as ad0, "the first > disk existing", as well as if an ATA disk would be present, > but no SATA disk. > > I can understand that this is annoying. > > > >> I don't >> care what it comes out as, as long as it stays put... > > There are labels or UFSIDs you can use to identify partitions > on a disk regardless of the device name they would come > out as. See "man glabel" and "man tunefs" for details. > > > >> Since I have >> the only fbsd system at work, I tend to format alot of funky drives >> for people, and it gets anoying having to swap fstab's every time... > > Labels can really help here: /etc/fstab will then contain > labels or UFSIDs instead of device names - and they don't > change when a disk is added or removed. > > > -- > Polytropon > Magdeburg, Germany > Happy FreeBSD user since 4.0 > Andra moi ennepe, Mousa, ... > I knew there had to be a solution! Thanks! Steve ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: can't make an 'a' slice except with auto-defaults
On Wed, Feb 03, 2010 at 06:37:22PM +0100, Polytropon wrote: > On Wed, 3 Feb 2010 10:26:06 -0700, Steve Franks > wrote: > > I'm just trying > > to make a fresh disk ready for dump/restore with a 1G /, so I guess > > sysinstall is out as an option at this time. > > Why so complicated? The command > > # newfs /dev/ad1 > > will prepare the disk, assuming ad1 is the new disk. It > will create one single partition covering the whole disk, > with no slice. You can then run > > # mount /dev/ad1 /mnt > > to access it; /dev/ad1 is the same as /dev/ad1c, "the > whole disk". This gets you what is referred to as a "dangerously dedicated" disk in the documentation.If you are doing nothing unusual with other OSen, then it works fine.But, it is also not at all hard to use fdisk and bsdlabel to create the full slice+partition. Here is the basic routine. dd if=/dev/zero of=/dev/ad1 bs=512 count=1024 fdisk -BI ad1 dd if=/dev/zero of=/dev/ad1s1 bs=512 count=1024 bsdlabel -w -B ad1s1 bsdlabel -e ad1s1 The second bsdlabel command puts you in to an edit session. Edit the partition table to something like this, then save[write] and exit. (Just those two lines for a single partition disk) # /dev/ad1s1: 8 partitions: #size offset fstype [fsize bsize bps/cpg] c: 89867610 0 unused0 0 # "raw" part, don't edit h:* 0 4.2BSD 2048 16384 8 You can use any partition identifier except 'c', but I like to reserve 'a' for bootable root (/) and 'b' for swap to reduce the confusion in my head. If the disk will not be bootable take out the 'B' from each of the fdisk and bsdlabel commands. You probably do not need the two dd-s, but sometimes they are needed if the system cannot read the sector 0 on the disk for some reason. Finally, when that is done, do: newfs /dev/ad1s1d Make yourself a mount point and mount the new disk. mkdir /bigwork mount /dev/ad1s1d /bigwork Edit /etc/fstab so it will mount automatically. It is not weird or mysterious, though I admit the man pages for both fdisk and bsdlabel could stand a going through. They do not follow the conventions of most man pages in the way they describe the switches and parameters. Fortunately they have lots of examples (but could use even more). jerry > > By the way, have you tried the program > > # sade > > instead of sysinstall? > > > -- > Polytropon > Magdeburg, Germany > Happy FreeBSD user since 4.0 > Andra moi ennepe, Mousa, ... > ___ > freebsd-questions@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-questions > To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org" ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: adding disk moves ad0 to ad4
Steve Franks wrote: Just curious, having read the handbook section talking about freebsd going straight to the hardware and skipping the bios for disk numbering, why then, if I stick a sata disk in 'sata0' on the motherboard, does it come up as ad0, but if I add a second disk in 'sata1' or 'pata0', on the next boot, I have no ad0, but ad4 and ad6? This seems to be the case with every mobo I've owned in the last 2 years from a variety of mfr's. Is there a way around this? I don't care what it comes out as, as long as it stays put... Since I have the only fbsd system at work, I tend to format alot of funky drives for people, and it gets anoying having to swap fstab's every time... If you think ad4 and ad6 for your disks is odd take a look at this (slightly line wrapped) extract from my file server boot dmesg. art...@fileserver> grep ata /var/run/dmesg.boot | grep -v ITHREAD atapci0: mem 0xf410-0xf4101fff irq 19 at device 0.0 on pci2 atapci0: AHCI Version 01.00 controller with 2 ports detected ata2: on atapci0 ata3: on atapci0 atapci1: port 0xb000-0xb007,0xb100-0xb103,0xb200-0xb207, 0xb300-0xb303,0xb400-0xb40f irq 16 at device 0.1 on pci2 ata4: on atapci1 atapci2: port 0xe700-0xe707,0xe800-0xe803,0xe900-0xe907, 0xea00-0xea03,0xeb00-0xeb1f mem 0xf4286000-0xf42867ff irq 19 at device 31.2 on pci0 atapci2: AHCI Version 01.20 controller with 6 ports detected ata5: on atapci2 ata6: on atapci2 ata7: on atapci2 ata8: on atapci2 ata9: on atapci2 ata10: on atapci2 ad10: 476940MB at ata5-master SATA300 ad12: 953869MB at ata6-master SATA300 ad14: 953869MB at ata7-master SATA300 ad16: 953869MB at ata8-master SATA300 acd0: DVDR at ata9-master SATA150 My SATA disks are ad{10,12,14,16}. This is a 2 year old Gigabyte server mobo, I can't remember the exact model off hand. ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: can't make an 'a' slice except with auto-defaults
On Wed, Feb 03, 2010 at 10:42:42AM -0700, Steve Franks wrote: > > Why so complicated? The command > > > > # newfs /dev/ad1 > > I'm looking to mirror/dup/image the entire system to something I can > stick in another system.I hear there's good reasons for not > running my whole system off of a single partition. The three main reasons are: Backups.It is often easier to manage backups when the disk storage is thoughtfully divided into reasonable and functional pieces. Emergencies. If your system crashes, especially if it is due to some disk problem, you may need to boot your system to single user. In that case you will start with only / (root) mounted as read-only. Having a fairly small root partition means the chance of having the bad disk area be in what you are trying to mount is reduced (not eliminated, of course). Anyway, you may be more able to get up to a minimal system and then work on recovering the other partitions. Boot time. A possible benefit is that only root needs to be fsck-ed before other things can start. Remaining fsck-s can run in parallel. This will take you less time to get back up after an abnormal shutdown - such as from a sudden power loss. > The 'other' system > has 7.2 and has devolved to a 25% chance of a hard freeze every time I > unplug a ucom device (seems to have cropped up between 7.2-release and > 7.2-stable#3). 8 likes usb, so I like 8. FreeBSd 8 is a good choice. jerry > > Steve > ___ > freebsd-questions@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-questions > To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org" > ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: kmem_size / arc_max ratio
In the last episode (Feb 03), Rolf Nielsen said: > I recently installed more RAM in my computer; increased it from 4GB to > 8GB. Mainly to be able to use a larger arc for zfs. Now my question is, > how much memory does the kernel need apart from the arc? I currently have > kmem_size_max 512MB bigger than arc_max. I haven't run into any problems > so far. I use my computer as desktop only, running WindowMaker and > normally having xconsole, emiclock, firefox with 5+ tabs, thunderbird, > amsn, sunbird, thunar, ktorrent and upto 5 or 6 xterms. > > My loader.conf looks like this: > > vfs.root.mountfrom="zfs:sysroot" > vm.kmem_size_max=5G > vm.kmem_size=5G > vfs.zfs.arc_max=4608M > vm.pmap.pg_ps_enabled=1 > zfs_load="YES" > nvidia_load="YES" vm.kmem_size_max should default to something like 300GB now, so you can remove that line. vm.pmap.pg_ps_enabled defaults to 1 nowadays, too. I have an 8GB system and use these settings: vm.kmem_size=6G vfs.zfs.arc_max=5G -- Dan Nelson dnel...@allantgroup.com ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: Backup and FreeBSD/ZFS
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 03.02.2010 19:14, LoH wrote: > If my memory serves, you're looking at something similar to taking a > snapshot and then sending it to the tape device, so zfs send > | (tape device access). This, IIRC, is functionally identical to > dump/restore. Except for one smallish detail. Dump handles "tape is full, switch to next one" in a relatively painless way... Let's just say that ... there's a reason I've invested in an autoloader for my home server (it will, among other things, hold about a terabyte of Nikon .NEF files if that means anything) //Svein - -- - +---+--- /"\ |Svein Skogen | sv...@d80.iso100.no \ / |Solberg Østli 9| PGP Key: 0xE5E76831 X|2020 Skedsmokorset | sv...@jernhuset.no / \ |Norway | PGP Key: 0xCE96CE13 | | sv...@stillbilde.net ascii | | PGP Key: 0x58CD33B6 ribbon |System Admin | svein-listm...@stillbilde.net Campaign|stillbilde.net | PGP Key: 0x22D494A4 +---+--- |msn messenger: | Mobile Phone: +47 907 03 575 |sv...@jernhuset.no | RIPE handle:SS16503-RIPE - +---+--- If you really are in a hurry, mail me at svein-mob...@stillbilde.net This mailbox goes directly to my cellphone and is checked even when I'm not in front of my computer. - Picture Gallery: https://gallery.stillbilde.net/v/svein/ - -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.12 (MingW32) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAktpzMgACgkQODUnwSLUlKRRfgCgulZAvQN61uE6HIcuvxzkU2yS HaAAmwcHY6YYqoTYlw/R/KeWuy/9ferH =zHrR -END PGP SIGNATURE- ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: Backup and FreeBSD/ZFS
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 03.02.2010 17:41, Goran Lowkrantz wrote: > I have been using Amanda straight from the port, it supports both tar > from snapshots if you need to be able to retrieve individual files from > the backup and zfs send if recovery at filesystem level is OK. > Does Amanda handle splitting a backup over several tapes (and using the autoloader under FreeBSD?) //Svein - -- - +---+--- /"\ |Svein Skogen | sv...@d80.iso100.no \ / |Solberg Østli 9| PGP Key: 0xE5E76831 X|2020 Skedsmokorset | sv...@jernhuset.no / \ |Norway | PGP Key: 0xCE96CE13 | | sv...@stillbilde.net ascii | | PGP Key: 0x58CD33B6 ribbon |System Admin | svein-listm...@stillbilde.net Campaign|stillbilde.net | PGP Key: 0x22D494A4 +---+--- |msn messenger: | Mobile Phone: +47 907 03 575 |sv...@jernhuset.no | RIPE handle:SS16503-RIPE - +---+--- If you really are in a hurry, mail me at svein-mob...@stillbilde.net This mailbox goes directly to my cellphone and is checked even when I'm not in front of my computer. - Picture Gallery: https://gallery.stillbilde.net/v/svein/ - -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.12 (MingW32) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAktpzQoACgkQODUnwSLUlKT1KwCgqZjUakGildXBWt4WRF/k6x5b NPwAn0uMvmseUwXHCpcxAu9uzdQfMhnJ =hAB3 -END PGP SIGNATURE- ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
IPv6: rtsol must be run a second time after boot to pick up default route
I recently set up an HE.net tunnel using the following guides: http://www.freebsd.org/doc/handbook/network-ipv6.html http://www.freebsddiary.org/ipv6.php FreeBSD 7.2-p5 is used for the router and the host, and it works beautifully, except that the host will only pick up the IPv6 prefix on boot and set its IP accordingly (local network functions), but will NOT set the default route unless I wait up to 10 minutes for the advertisement, or manually run rtsol. The same problem happens with OS X 10.6.2, but not with Win7 (and Linux 2.6 remains untested at this time). The host has no firewall running currently, and there's no firewalling between the router and the host. Running rtsol with debugging doesn't show anything out of the ordinary, either during boot or after. Rtadvd is running on the router and my setup is identical to the guides other than device name: $ cat /etc/rtadvd.conf vr1:\ :addrs#1:addr="2001:470::::":prefixlen#64:tc=ether: Any suggestions? I've tried a few variations of rtadvd.conf without any changes in behavior. I'm inclined to think it's router-related, given the issue on multiple OSes, but I suppose it could go either way. I'd much prefer not to add in extra calls of rtsol in /etc/rc.local. Thanks. Brian Conway ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: Backup and FreeBSD/ZFS
Yes, I have three sets running that way, my home systems using a Dell 122 and a 5 rack config at work with a 7x200G Tandberg and a 48 slot 4U IBM beast. All are using the mtx changer scripts that come with Amanda. -glz --On Wednesday, February 03, 2010 8:22 PM +0100 "Svein Skogen (Listmail Account)" wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On 03.02.2010 17:41, Goran Lowkrantz wrote: I have been using Amanda straight from the port, it supports both tar from snapshots if you need to be able to retrieve individual files from the backup and zfs send if recovery at filesystem level is OK. Does Amanda handle splitting a backup over several tapes (and using the autoloader under FreeBSD?) //Svein - -- - +---+--- /"\ |Svein Skogen | sv...@d80.iso100.no \ / |Solberg Østli 9| PGP Key: 0xE5E76831 X|2020 Skedsmokorset | sv...@jernhuset.no / \ |Norway | PGP Key: 0xCE96CE13 | | sv...@stillbilde.net ascii | | PGP Key: 0x58CD33B6 ribbon |System Admin | svein-listm...@stillbilde.net Campaign|stillbilde.net | PGP Key: 0x22D494A4 +---+--- |msn messenger: | Mobile Phone: +47 907 03 575 |sv...@jernhuset.no | RIPE handle:SS16503-RIPE - +---+--- If you really are in a hurry, mail me at svein-mob...@stillbilde.net This mailbox goes directly to my cellphone and is checked even when I'm not in front of my computer. - Picture Gallery: https://gallery.stillbilde.net/v/svein/ - -BEGIN PGP SIGNATURE- Version: GnuPG v2.0.12 (MingW32) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAktpzQoACgkQODUnwSLUlKT1KwCgqZjUakGildXBWt4WRF/k6x5b NPwAn0uMvmseUwXHCpcxAu9uzdQfMhnJ =hAB3 -END PGP SIGNATURE- ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org" ... the future isMobile Goran Lowkrantz System Architect, isMobile, Aurorum 2, S-977 75 Luleå, Sweden Phone: +46(0)920-75559 Mobile: +46(0)70-587 87 82 Fax: +46(0)70-615 87 82 http://www.ismobile.com ... ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
More sysinstall questions 1 of 2
Yesterday, I asked how sysinstall mounts the drive on which FreeBSD is to install. I might not have been clear enough so I will try again since my question may have been confusing. The system is booting via mfs so we are starting out with a virtual disk drive made of memory. The hard drive is sitting right there as /dev/ad0. It can be formatted and mounted and appears to be working properly. As a trouble-shooting step, I ran sysinstall from mfs manually exactly as I have done from a CDROM on that very box. With the mfs system, sysinstall sees the hard drive and appears to let you format it. The bsdlabel section appears to let you assign the partitions. One selects distributions and a ftp site and then . . . it all goes wrong. The commit does not format the disk. There is no "last chance" prompt. It goes right to the download and proceeds to install FreeBSD all over mfs. The CDROM for installing FreeBSD correctly formats the drive and installs the OS on that very same box. My first question is why doesn't the mfs do the same thing? My second question will be on a separate message. Martin McCormick ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
More sysinstall questions 2 of 2
When looking at the screen in sysinstall that lets one choose a medium to import an install.cfg file, there is the CDROM, the floppy disk and what looks like another option. Here is the text of the screen: ³ ³ fd0floppy drive unit A³ ³ ³ ³ acd0 ATAPI/IDE CDROM³ ³ ³ ³ ufsid/4b6787a7598a9a8bs1a ufsid/4b6787a7598a9a8bs1a ³ ³ Is that line with the ufsid some way to import a file without having to install a CDROM or some other physical media? Thanks for your help. Martin McCormick WB5AGZ Stillwater, OK Systems Engineer OSU Information Technology Department Telecommunications Services Group ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: More sysinstall questions 1 of 2
On Wed, 03 Feb 2010 17:01:52 -0600, Martin McCormick wrote: > The commit does not format the disk. There is no "last > chance" prompt. It goes right to the download and proceeds to > install FreeBSD all over mfs. This seems to be obvious because no changes have been made to the disk (i. e. no slicing, no MBRin, no partitioning). > The CDROM for installing FreeBSD correctly formats the > drive and installs the OS on that very same box. My first > question is why doesn't the mfs do the same thing? This is a matter of what has been selected within the partition editor. If not "UFS2+S Y" is set, no formatting process will take place. -- Polytropon Magdeburg, Germany Happy FreeBSD user since 4.0 Andra moi ennepe, Mousa, ... ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: More sysinstall questions 2 of 2
On Wed, 03 Feb 2010 17:09:28 -0600, Martin McCormick wrote: >³ ³ ufsid/4b6787a7598a9a8bs1a ufsid/4b6787a7598a9a8bs1a ³ ³ > > Is that line with the ufsid some way to import a file without > having to install a CDROM or some other physical media? To me, it seems to refer to an obviously UFS formatted media; maybe this is the hard disk (which has a UFS partition on it accessible via its UFSID)? The last part "s1a" may suggest that it is s1a (of ad0)... -- Polytropon Magdeburg, Germany Happy FreeBSD user since 4.0 Andra moi ennepe, Mousa, ... ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
this may be impossible: iis there a way to play streams on our firefox?
Is there a way of streaming stuff like old tv shows, the bbc-player, and pbs steams on the freebsd version of firefox35? thought i'd give it a last try since everything is upgraded on my desktop. -- Gary Kline kl...@thought.org http://www.thought.org Public Service Unix http://jottings.thought.org http://transfinite.thought.org The 7.79a release of Jottings: http://jottings.thought.org/index.php ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: IPv6: rtsol must be run a second time after boot to pick up default route
On Wed, 3 Feb 2010, Brian Conway wrote: I recently set up an HE.net tunnel using the following guides: http://www.freebsd.org/doc/handbook/network-ipv6.html http://www.freebsddiary.org/ipv6.php FreeBSD 7.2-p5 is used for the router and the host, and it works beautifully, except that the host will only pick up the IPv6 prefix on boot and set its IP accordingly (local network functions), but will NOT set the default route unless I wait up to 10 minutes for the advertisement, or manually run rtsol. The same problem happens with OS X 10.6.2, but not with Win7 (and Linux 2.6 remains untested at this time). The host has no firewall running currently, and there's no firewalling between the router and the host. Running rtsol with debugging doesn't show anything out of the ordinary, either during boot or after. Rtadvd is running on the router and my setup is identical to the guides other than device name: $ cat /etc/rtadvd.conf vr1:\ :addrs#1:addr="2001:470::::":prefixlen#64:tc=ether: Any suggestions? I've tried a few variations of rtadvd.conf without any changes in behavior. I'm inclined to think it's router-related, given the issue on multiple OSes, but I suppose it could go either way. I'd much prefer not to add in extra calls of rtsol in /etc/rc.local. Thanks. Brian Conway A few more (unusual) details as follow-up: - The missing route doesn't happen on Win7 or Linux 2.6 (Debian 5.0/Lenny) - The missing route still happens on both OS X 10.6.2 and FreeBSD 7.2-p5 - This ONLY happens after a warm reboot. Neither FreeBSD nor OS X have the issue with a cold boot. The boot-up's rtsol picks up the default route immediately. Weird. Brian Conway ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: this may be impossible: iis there a way to play streams on our firefox?
On Wed, Feb 3, 2010 at 7:29 PM, Gary Kline wrote: > > >Is there a way of streaming stuff like old tv shows, the >bbc-player, and pbs steams on the freebsd version of firefox35? > >thought i'd give it a last try since everything is upgraded on >my desktop. > /usr/ports/www/linux-f10-flashplugin10 -- Adam Vande More ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: GeForce GTX 260M on amd64
Hello, I found this - now I am waiting if this will be in ports: http://www.nvnews.net/vbulletin/showthread.php?t=142120 Radek ___ freebsd-questions@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-questions To unsubscribe, send any mail to "freebsd-questions-unsubscr...@freebsd.org"
Re: can't make an 'a' slice except with auto-defaults
In freebsd-questions Digest, Vol 296, Issue 6, Message: 20 On Wed, 3 Feb 2010 10:26:06 -0700 Steve Franks wrote: > On Wed, Feb 3, 2010 at 8:18 AM, Jerry McAllister wrote: > > On Tue, Feb 02, 2010 at 07:59:15PM -0700, Steve Franks wrote: > > > >> On a running system. I mean, I know I should quit being a &%^#& and > >> read the manpage for bsdlabel, but sysintall really does have a nice > >> tui. 'C'reate slice goes straight to 'd', even on a 'fresh' disk. sysinstall (or sade) will assign 'd' to the first partition if its mount point for the partition is not specified as '/'. This is usually right, when partitioning either another slice on the same disk, or a slice on another disk, where '/' is already assigned to the booted disk. In the case of what you're doing, ie preparing another disk to copy your system to, it's not hard to fix that later with bsdlabel -e, as Jerry has pointed out, and which I'll detail further below. > >> I see in the handbook, this is alluded to, but some intermediate level > >> between begginer and expert (bsdlabel just strikes me as way too easy > >> to trash the disk I'm running off of while trying to make a backup), > >> would be nice...512M just won't fit the kernel+symbols. A default 1GB '/' is on the cards, perhaps by 8.1, but in any case I've never used anything like the autodefaults for other partitions either. bsdlabel(8) is not so scary. For one thing, anything you do is shown but not committed to disk if you use the -n switch. Try it. Also, the man has examples of saving and restoring an existing label, so if you: # bsdlabel ad1s1 > saved.ad1s1.label then if you stuff it up you can later on just restore it with: # bsdlabel -R ad1s1 saved.ad1s1.label > >> > > > > Well, Create slice would be an fdisk(8) thing, not bsdlabel. > > bsdlabel creates partitions within a slice. > > > > But, generally you cannot run fdisk on a disk that is in use on a > > running system - which generally means that it is the boot device, > > has filesystems mounted or has part of the currently designated swap > > space. You will need to plug in a boot cd or bring up the fixit system > > for that. The fixit system runs from memory - creates filesystems > > and mount points in memory rather than on disk, so it can talk to > > any disk. If you set kern.geom.debugflags=16 and 'w'rite from either sysinstall's fdisk or bsdlabel screens you can update the partition table or your new slice's bsdlabel, but you have to be very careful, and in the case of fdisk, you need to reboot before labeling the new slice. Certainly using a fixit boot is the safe and sure way to avoid complications. > > New, if you are working on a non-used (extra) disk, eg one that is not > > the boot device nor has any mounted filesystems or swap space > > on it, then you should be able to fdisk and bsdlabel that from > > a running system. This seems to be Steve's case, but he's right; it will start creating new partitions as 'd' rather than 'a' (since there's already a '/') unless he boots into either sysinstall or fixit from another source. > > I have no idea what you mean by "'C'reate slice goes straight to 'd'" > > It does not match anything I remember being possible. I don't happen > > to have any system handy at the moment that I can muck with disks on. > > > > jerry I've several times added partitions to extra slice/s on either the boot disk or added disks (including sliced USB flash disks) using sysinstall invoked from the running system, and these do start with 'd' partition. > Ok, terminology crash. As someone pointed out, I'm talking about > label, here, not fdisk, and partitions, not slices (had those two > backwards in my head). > > Basically, as far as I can tell, on a running system, there is no > combination of keystrokes in sysinstall's label editor that will > create an "ad[1-9]s1a", except the 'a' key which produces a 512M s1a. The 'a' key auto-assigns '/' as the mount point for partition 'a', which is why you see that. You wouldn't be able to commit that anyway, as /, /var, /usr would conflict with your already mounted slice, and newfs'ing your existing system is most likely not what you want :) > All other keystrokes (namely 'c') go straight to "ad[1-9]s1d" when a > second disk is placed in a system booted from ad0s1a. I'm just trying > to make a fresh disk ready for dump/restore with a 1G /, so I guess > sysinstall is out as an option at this time. You may be better off just installing the new system onto ad1 straight up, ignoring your ad0, when you can just use sysinstall. However .. sysinstall (or sade) run from an existing system is a pretty convenient way to partition a disk, or slice. You don't really need to worry about it starting at 'd', as you can easily correct that later. Eg this one: smithi on sola% fdisk -s ad0 /dev/ad0: 77520 cyl 16 hd 63 sec PartStartSize Type Flags 1: