Re: Advanced Format Drive ?
In message , Warren Block wrote: >On Wed, 28 Nov 2012, Ronald F. Guilmette wrote: > >> >> In message , >> Warren Block wrote: >> I tried to do as you suggest and change the partition type to freebsd-ufs, but there's a problem... # gpart modify -i 1 -t freebsd-ufs /dev/da1 gpart: Invalid argument >>> >>> da1 is the drive. da1s1 is the first slice. >> >> Yeabut that's what I thought that the -i option was for! I mean isn't the >> parameter for >> that supposed to tell gpart which sub-part of the whole "geom" thing that is >> named on >> the command line is supposed to be modified? > >Doh, you are right, and I misread it. It's the type that is the >problem: freebsd-ufs is a not an MBR partition type, it should be just >freebsd: > ># gpart modify -i1 -t freebsd /dev/da1 Thank you again Warren. Success! I am a happy camper! % gpart show /dev/da1 =>63 1953525104 da1 MBR (931G) 631985 - free - (992k) 2048 19535196161 freebsd (931G) 19535216643503 - free - (1.7M) ___ 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: Advanced Format Drive ?
On Wed, 28 Nov 2012, Ronald F. Guilmette wrote: In message , Warren Block wrote: I tried to do as you suggest and change the partition type to freebsd-ufs, but there's a problem... # gpart modify -i 1 -t freebsd-ufs /dev/da1 gpart: Invalid argument da1 is the drive. da1s1 is the first slice. Yeabut that's what I thought that the -i option was for! I mean isn't the parameter for that supposed to tell gpart which sub-part of the whole "geom" thing that is named on the command line is supposed to be modified? Doh, you are right, and I misread it. It's the type that is the problem: freebsd-ufs is a not an MBR partition type, it should be just freebsd: # gpart modify -i1 -t freebsd /dev/da1 ___ 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: Advanced Format Drive ?
In message , Warren Block wrote: >> I tried to do as you suggest and change the partition type to freebsd-ufs, >> but there's a problem... >> >> # gpart modify -i 1 -t freebsd-ufs /dev/da1 >> gpart: Invalid argument > >da1 is the drive. da1s1 is the first slice. Yeabut that's what I thought that the -i option was for! I mean isn't the parameter for that supposed to tell gpart which sub-part of the whole "geom" thing that is named on the command line is supposed to be modified? Well, now I'm totally confused, but I'll try it again in the way I think you are suggesting... # gpart modify -i 1 -t freebsd-ufs /dev/da1s1 gpart: geom '/dev/da1s1': Invalid argument U... What else should I try? ___ 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: Advanced Format Drive ?
On Tue, 27 Nov 2012, Ronald F. Guilmette wrote: In message , Warren Block wrote: On Mon, 26 Nov 2012, Ronald F. Guilmette wrote: Starting sector 2048 is definitely a multiple of 4KB, so I am assuming that all I really need to do here in order to use this new drive as extra stroage for a FreeBSD system (assuming that I am happy with good old fashioned MBR style partitioning, which I am) is just: newfs -U /dev/da1s1 You should also change the partition type to freebsd or freebsd-ufs. Offhand I'd guess they're both 0xa5, but have not looked. Use 'gpart modify'. Thanks Warren! I confess that I hadn't even thought about that. And as a result, the partition that I just backed up a substantial part of my system onto is still being listed as "ntfs", even though I have done the newfs to it and (thus) it is now actually a UFS partition... not NTFS. % gpart show /dev/da1 =>63 1953525104 da1 MBR (931G) 631985 - free - (992k) 2048 19535196161 ntfs (931G) 19535216643503 - free - (1.7M) I can still mount it as a UFS, no problem, so Im inclined to wonder what the type code on a partition is used for anyway. (FreeBSD doesn't seem to care if a partition is marked as NTFS as long as it actually has a UFS filesystem in it.) Mostly relevant when booting from that drive. Still, it would be bad for some NTFS utility to helpfully attempt repair of a UFS filesystem. I tried to do as you suggest and change the partition type to freebsd-ufs, but there's a problem... # gpart modify -i 1 -t freebsd-ufs /dev/da1 gpart: Invalid argument da1 is the drive. da1s1 is the first slice. (The error message "Invalid argument" is not terribly informative. It doesn't even indicate which argument is to blame. And I'm not sure if the index numbers that gpart uses start from 0 or from 1. The man page doesn't say.) Slice/partition number is the third column in the gpart output above. MBR slice numbering starts at one. P.S. When doing the newfs, I actually ended up having to do: newfs -U -f 4096 /dev/da1s1 because I was doing this on an old 8.3 system, so the default frag size there was still set at 2048. ___ 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: Advanced Format Drive ?
In message , Warren Block wrote: >On Mon, 26 Nov 2012, Ronald F. Guilmette wrote: > >> Starting sector 2048 is definitely a multiple of 4KB, so I am assuming >> that all I really need to do here in order to use this new drive as extra >> stroage for a FreeBSD system (assuming that I am happy with good old >> fashioned MBR style partitioning, which I am) is just: >> >>newfs -U /dev/da1s1 > >You should also change the partition type to freebsd or freebsd-ufs. >Offhand I'd guess they're both 0xa5, but have not looked. Use 'gpart >modify'. Thanks Warren! I confess that I hadn't even thought about that. And as a result, the partition that I just backed up a substantial part of my system onto is still being listed as "ntfs", even though I have done the newfs to it and (thus) it is now actually a UFS partition... not NTFS. % gpart show /dev/da1 =>63 1953525104 da1 MBR (931G) 631985 - free - (992k) 2048 19535196161 ntfs (931G) 19535216643503 - free - (1.7M) I can still mount it as a UFS, no problem, so Im inclined to wonder what the type code on a partition is used for anyway. (FreeBSD doesn't seem to care if a partition is marked as NTFS as long as it actually has a UFS filesystem in it.) I tried to do as you suggest and change the partition type to freebsd-ufs, but there's a problem... # gpart modify -i 1 -t freebsd-ufs /dev/da1 gpart: Invalid argument # gpart modify -i 0 -t freebsd-ufs /dev/da1 gpart: index '0': No such file or directory Obviously, I'm doing this wrong, but what is the Right Way? (The error message "Invalid argument" is not terribly informative. It doesn't even indicate which argument is to blame. And I'm not sure if the index numbers that gpart uses start from 0 or from 1. The man page doesn't say.) P.S. When doing the newfs, I actually ended up having to do: newfs -U -f 4096 /dev/da1s1 because I was doing this on an old 8.3 system, so the default frag size there was still set at 2048. ___ 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: Advanced Format Drive ?
On Mon, 26 Nov 2012, Ronald F. Guilmette wrote: Starting sector 2048 is definitely a multiple of 4KB, so I am assuming that all I really need to do here in order to use this new drive as extra stroage for a FreeBSD system (assuming that I am happy with good old fashioned MBR style partitioning, which I am) is just: newfs -U /dev/da1s1 You should also change the partition type to freebsd or freebsd-ufs. Offhand I'd guess they're both 0xa5, but have not looked. Use 'gpart modify'. ___ 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: Advanced Format Drive ?
On Mon, 26 Nov 2012, Ronald F. Guilmette wrote: I can't help but note that both this one and also my new Seagate 2.5" portable external 1TB drive are pre-partitioned with one partition that in both cases starts at sector 2048. I suppose that it is no coincidence that 2048 * 512B == 1 megabyte, exactly. Told ya. ___ 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"
Advanced Format Drive ?
I just wanted to add another data point... in case anybody is interested... uring my recent spending binge, I also acquired a Hatachi portable external 2.5 inch "Touro Moble" 500GB drive. Righ out of the box, this is what fdisk tells me about this one: == *** Working on device /dev/da1 *** parameters extracted from in-core disklabel are: cylinders=60801 heads=255 sectors/track=63 (16065 blks/cyl) Figures below won't work with BIOS for partitions not in cyl 1 parameters to be used for BIOS calculations are: cylinders=60801 heads=255 sectors/track=63 (16065 blks/cyl) Media sector size is 512 Warning: BIOS sector numbering starts with sector 1 Information from DOS bootblock is: The data for partition 1 is: sysid 7 (0x07),(NTFS, OS/2 HPFS, QNX-2 (16 bit) or Advanced UNIX) start 2048, size 976766976 (476937 Meg), flag 0 beg: cyl 0/ head 32/ sector 33; end: cyl 1023/ head 254/ sector 63 The data for partition 2 is: The data for partition 3 is: The data for partition 4 is: == I can't help but note that both this one and also my new Seagate 2.5" portable external 1TB drive are pre-partitioned with one partition that in both cases starts at sector 2048. I suppose that it is no coincidence that 2048 * 512B == 1 megabyte, exactly. ___ 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"
Advanced Format Drive ?
Just a brief follow-up on my questions here (about 2 weeks ago) regarding so-called "Advanced Format" (4KB block) drives... I just got myself a shiny new Seagate 2.5" portable external 1TB hard drive. fdisk is telling me this about it: === *** Working on device /dev/da1 *** parameters extracted from in-core disklabel are: cylinders=121601 heads=255 sectors/track=63 (16065 blks/cyl) Figures below won't work with BIOS for partitions not in cyl 1 parameters to be used for BIOS calculations are: cylinders=121601 heads=255 sectors/track=63 (16065 blks/cyl) Media sector size is 512 Warning: BIOS sector numbering starts with sector 1 Information from DOS bootblock is: The data for partition 1 is: sysid 7 (0x07),(NTFS, OS/2 HPFS, QNX-2 (16 bit) or Advanced UNIX) start 2048, size 1953519616 (953867 Meg), flag 0 beg: cyl 0/ head 32/ sector 33; end: cyl 1023/ head 254/ sector 63 The data for partition 2 is: The data for partition 3 is: The data for partition 4 is: === Starting sector 2048 is definitely a multiple of 4KB, so I am assuming that all I really need to do here in order to use this new drive as extra stroage for a FreeBSD system (assuming that I am happy with good old fashioned MBR style partitioning, which I am) is just: newfs -U /dev/da1s1 Right? Last question: How could I even tell if this thing is or isn't "Advanced Format"? Is there some tool I could run that would tell me that? ___ 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: Advanced Format Drive ? GPT ?
Warren Block wrote: On Thu, 15 Nov 2012, Al Plant wrote: I looked over the GPT sample and have a question. In the fstab entries, something that uses msdosfs, (thumb drive maybe). %%%>> Can you enter it directly in the fstab after the basic partitions and other /dev have been entered in the initial setup? Short answer: yes, but... Longer answer: most flash drives have an MBR partition setup with one partition filling the whole device. Since it's not GPT, it won't/can't have GPT labels on the partitions. But the GEOM system will create a label for the MSDOS filesystem if it has been given a volume name. That label will appear in /dev/msdosfs/ and can be used in an /etc/fstab entry. ___ Thanks,, For the sage advice. % ~Al Plant - Honolulu, Hawaii - Phone: 808-284-2740 + http://hawaiidakine.com + http://freebsdinfo.org + + http://aloha50.net - Supporting - FreeBSD 7.2 - 8.0 - 9* + < email: n...@hdk5.net > "All that's really worth doing is what we do for others."- Lewis Carrol ___ 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: Advanced Format Drive ?
On 11/16/12 14:07, dweimer wrote: On 2012-11-15 17:31, Ronald F. Guilmette wrote: (This stuff would probably be a lot less confiusing if I actually knew what I was doing, but...) OK, Warren, I've just done the following steps. The first two I drew from the manpage examples, and then followed those up with two commands from your tutorial. /sbin/gpart create -s GPT ada0# manpage example is wrong, ad0 -> ada0 /sbin/gpart bootcode -b /boot/mbr ad0 # manpage wrong again, pmbr -> mbr gpart add -t freebsd-boot -l gpboot -b 40 -s 512K ada0 gpart bootcode -b /boot/pmbr -p /boot/gptboot -i 1 ada0 That last one, done at the suggestion of your tutorial page, has me completely perplexed, because of what is said, very explicitly, in the gpart(8) manpage: bootcode Embed bootstrap code into the partitioning scheme's metadata on the geom (using -b bootcode) or write bootstrap code into a partition (using -p partcode and -i index). Please note the use of the word "or". The man page is telling me to _either_ use the -p option _or else_ use the -p and -i options together. But you are telling me to use all three in one go! Forgive me, but I'm confused. (As you can tell by now, I am often easily confused. Sorry.) ___ 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" I saw this, and well started wondering myself, as I have been using this while doing work on booting FreeBSD via ZFS (of course using -p /boot/gptzfsboot), I got the line from a tutorial on booting from ZFS. Never thought much of it, until now, but I believe I see now why, the secret is the pmbr, notice the "p". Its the protective mbr, it lets formatting tools that understand mbr, but not gpart know that there is something there, the actual boot code is in the partition. pmbr serves two purposes. It's both the first stage boot code, as a traditional BIOS always loads the first block of the disk into memory and runs it to boot the machine regardless of whether you've got an MBR or GPT disk, and it contains a traditional MBR that shows the entire disk is occupied by the first DOS partition (slice in BSD terminology) and that is of type 0xee. The latter means that GPT ignorant utilities see the disk as fully occupied by a partition of unknown type, which should mean they won't touch anything. The pmbr boot code understands the GPT table and runs through the partition entries looking for one of type freebsd-boot. When it finds one, it then loads the contents of the partition (or the first 545k if it's larger) into memory and jumps to the second stage boot loader. ___ 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: Advanced Format Drive ?
On 2012-11-15 17:31, Ronald F. Guilmette wrote: (This stuff would probably be a lot less confiusing if I actually knew what I was doing, but...) OK, Warren, I've just done the following steps. The first two I drew from the manpage examples, and then followed those up with two commands from your tutorial. /sbin/gpart create -s GPT ada0# manpage example is wrong, ad0 -> ada0 /sbin/gpart bootcode -b /boot/mbr ad0 # manpage wrong again, pmbr -> mbr gpart add -t freebsd-boot -l gpboot -b 40 -s 512K ada0 gpart bootcode -b /boot/pmbr -p /boot/gptboot -i 1 ada0 That last one, done at the suggestion of your tutorial page, has me completely perplexed, because of what is said, very explicitly, in the gpart(8) manpage: bootcode Embed bootstrap code into the partitioning scheme's metadata on the geom (using -b bootcode) or write bootstrap code into a partition (using -p partcode and -i index). Please note the use of the word "or". The man page is telling me to _either_ use the -p option _or else_ use the -p and -i options together. But you are telling me to use all three in one go! Forgive me, but I'm confused. (As you can tell by now, I am often easily confused. Sorry.) ___ 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" I saw this, and well started wondering myself, as I have been using this while doing work on booting FreeBSD via ZFS (of course using -p /boot/gptzfsboot), I got the line from a tutorial on booting from ZFS. Never thought much of it, until now, but I believe I see now why, the secret is the pmbr, notice the "p". Its the protective mbr, it lets formatting tools that understand mbr, but not gpart know that there is something there, the actual boot code is in the partition. -- Thanks, Dean E. Weimer http://www.dweimer.net/ ___ 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: Advanced Format Drive ?
On Thu, 15 Nov 2012, Ronald F. Guilmette wrote: I think that I have only two final questions: 1) I can't remember now if the ``guided'' partitioning approach that is offered to folks who are installing FreeBSD 9.x itself offers a "GPT" option or not. Does it? (If not, and if MBR is really now considered antiquated, then I would think that the install process really should offer a GPT option, if it isn't doing so already.) GPT is the default for bsdinstall. 2) Not knowing any better, on this fresh install that I'm doing now (of 9.1-RC3) when it got down to the point where it asked me how I wanted to partition, I selected the "exit to shell" option. Once I got a shell prompt, I proceeded to do bascially everything that's suggested in the "The New Standard Method" section of Warren's nice tutorial. My assumption was that I could do this, get all of my shiny new GPT partitions just the way I wanted them, and just simply exit the shell... an action which, I had hoped, would return me to the install process at a point where I would then be asked to assign mount points to each of my newly created GPT partitions, and then, hopefully, the rest of the install process would proceed in an entirely customary way. It would, but you have to mount the new filesystems in a certain spot. bsdinstall shows a prompt about that. And how exactly do mount points get associated with partitions (in particular GPT partitions) anyway? Are these just another partition attribute? The gpart(8) man page is also utterly silent on the subject of mount points, even though they are quite obviously a rather critical component of what it takes to make a partition useful on/to FreeBSD. GPT partitions appear in /dev as the drive name followed by "p" and the partition number, similar to the old slice/partition notation. So instead of /dev/ada0s1a, it will typically be /dev/ada0p2. These are entered in /etc/fstab as normal. My guide uses GPT labels, which are superior in many ways to fixed device names, but also not really covered by that article. P.S. Assigning mount points appears to be one thing that the new swiss- army-knife of gpart _cannot_ do. Given that, I have to ask... What if any command line tool is available to associate partitions with mount points? /etc/fstab, same as normal. ___ 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: Advanced Format Drive ?
NEVERMIND! It took me awhile, but I think I've finally got the hang of this gpart/GPT stuff... well... mostly anyway (but see below). I understand now that /boot/mbr is a "regular" sort of MBR, with regular sort of MBR bootstrap code, whereas /boot/pbmr is the ``protected'' MBR record that says, in effect "I am not a normal MBR. I am special and I'm here to tell you that this drive actually uses the GPT scheme." I've also figured out that despite the gpart(8) man page's unfortunate use of the word "or", it is evidently the case that when doing "gpart bootcode ..." it is perfectly OK to use all three of the -b, -p, and -i options together, and that doing so has (I think) the equivalent effect to invoking "gpart bootcode ..." once with only the -b option and then once again with only the -p and -i options together. I think that I have only two final questions: 1) I can't remember now if the ``guided'' partitioning approach that is offered to folks who are installing FreeBSD 9.x itself offers a "GPT" option or not. Does it? (If not, and if MBR is really now considered antiquated, then I would think that the install process really should offer a GPT option, if it isn't doing so already.) 2) Not knowing any better, on this fresh install that I'm doing now (of 9.1-RC3) when it got down to the point where it asked me how I wanted to partition, I selected the "exit to shell" option. Once I got a shell prompt, I proceeded to do bascially everything that's suggested in the "The New Standard Method" section of Warren's nice tutorial. My assumption was that I could do this, get all of my shiny new GPT partitions just the way I wanted them, and just simply exit the shell... an action which, I had hoped, would return me to the install process at a point where I would then be asked to assign mount points to each of my newly created GPT partitions, and then, hopefully, the rest of the install process would proceed in an entirely customary way. Sadly, this did not happen. After exiting from the shell, the install process _did_ resume, however the first thing it did was to check the integrity of the distributions (kernel+base) that I had selected earlier, and once it was satisfied that they were OK, it immediately started to try to extract everything from those two distribution files. It is easy to undeerstand why this last step failed virtually immediately with the error message: Error while extracting base.txz: Can't set user=0/group=0 for .Can't update time for . Obviously (and quite reasonably) the install process did not have any clear idea of where exactly it was supposed to be extracting the files to, because I had not even assigned mount points for any of my brand new GPT partitions yet. So, um, I'm wondering... Is this a bug, or a feature of the current FreeBSD install process? Should I be filing a PR on this? -=-=-=-=-=-=-=-=-=-=-=-=- A footnote: Since my first try at installing using a GPT partitioning scheme crashed and burned (as described above), I naturally hit the reset button on the machine in question and just started over from the beginning of the whole install process, hoping that I would (this time) be able to make use of the various GPT partitions that I had already set up. (See above.) After the obligatory preliminary questions, I finally I came to the place in the process where it wanted to know if I wanted to do guided, manual, or shell partitioning. At this point, I figured that it made little difference which one I choose (because I'd already created the partitions, and even newfs'd them all) and now I only needed to assign mount points. So I selected the easy choice... "guided"... and immediately I got an error message that said there is not enough free space on the drive to install FreeBSD. (Gr!) Oh well! Along with that error message, I was given an option to open the partition editor, so I took that option and then just assigned proper mount points to all of the partitions that I'd already created, and then, finally, I clicked on . >From there, everything went fine and I successfully installed a minimal 9.1-RC3 system, and then successfully booted it. But then I started to wonder if maybe Warren had left out the instructions for performing this final critical step (assigning mount points) in his tutorial, so I went back and looked at it to see if he had mentioned it, and I see that it doesn't say a word about mount points. Warren? Was this a deliberate or inadvertant ommission? Is the subject of mount points outside of the scope of what you had been intending to cover in your tutorial? And how exactly do mount points get associated with partitions (in particular GPT partitions) anyway? Are these just another partition attribute? The gpart(8) man page is also utterly silent on the subject of mount points, even though they are quite obviously a rather critical component of what it takes to make a partition useful on/to FreeBSD. P.S
Re: Advanced Format Drive ?
(This stuff would probably be a lot less confiusing if I actually knew what I was doing, but...) OK, Warren, I've just done the following steps. The first two I drew from the manpage examples, and then followed those up with two commands from your tutorial. /sbin/gpart create -s GPT ada0# manpage example is wrong, ad0 -> ada0 /sbin/gpart bootcode -b /boot/mbr ad0 # manpage wrong again, pmbr -> mbr gpart add -t freebsd-boot -l gpboot -b 40 -s 512K ada0 gpart bootcode -b /boot/pmbr -p /boot/gptboot -i 1 ada0 That last one, done at the suggestion of your tutorial page, has me completely perplexed, because of what is said, very explicitly, in the gpart(8) manpage: bootcode Embed bootstrap code into the partitioning scheme's metadata on the geom (using -b bootcode) or write bootstrap code into a partition (using -p partcode and -i index). Please note the use of the word "or". The man page is telling me to _either_ use the -p option _or else_ use the -p and -i options together. But you are telling me to use all three in one go! Forgive me, but I'm confused. (As you can tell by now, I am often easily confused. Sorry.) ___ 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: Advanced Format Drive ? GPT ?
On Thu, 15 Nov 2012, Al Plant wrote: I looked over the GPT sample and have a question. In the fstab entries, something that uses msdosfs, (thumb drive maybe). Can you enter it directly in the fstab after the basic partitions and other /dev have been entered in the initial setup? Short answer: yes, but... Longer answer: most flash drives have an MBR partition setup with one partition filling the whole device. Since it's not GPT, it won't/can't have GPT labels on the partitions. But the GEOM system will create a label for the MSDOS filesystem if it has been given a volume name. That label will appear in /dev/msdosfs/ and can be used in an /etc/fstab entry. ___ 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: Advanced Format Drive ?
Warren, In the EXAMPLES section of the gpart(8) man page, they do this: /sbin/gpart bootcode -b /boot/pmbr ad0 In your document however, you first create an explicit (special) partition named "gpboot" and then you do this instead: gpart bootcode -b /boot/pmbr -p /boot/gptboot -i 1 da0 Who is "right" in this case? I did the former, and did not get any error. The gpart(8) man page says: "... First, a protective MBR is embedded into the first disk sector from the /boot/pmbr image..." however it appears to me that the steps in your tutorial are effectively installing a copy of the /boot/pmbr file into block #40 of the disk. Yes? But isn't a copy of /boot/pmbr really supposed to end up in the first 512 bytes of the disk, i.e. block #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: Advanced Format Drive ?
In message , Warren Block wrote: >> It wouldn't hurt to add the above info to your tutorial page. > >The problem with that sort of detail is that too much of it obscures the >point, which in this case is just trying to show the right way to set up >disks without overwhelming the reader. Some things are inherently overwhelming, and there is no use to try to hide the fact. It just makes matters worse. >> P.S. I really do think that it is a serious omission that the gpart(8) >> man page doesn't really say anything regarding proper or desirable use >> of the "add -a" option. If it were up to me, I'd put in just a couple >> of short notes that would say at least something about 4K being Good and >> Desirable for modern drives, and 1M being Good and Desirable for SSDs. >> >> But maybe that kind of info does really belong in something more like >> an actual tutorial document... you know... something like, um, your's. > >I can see it both ways. A short mention of those values in that section >of gpart(8) would be helpful. The 1M value is controversial to some >people. Yeabut for some people, even evolution is "controversial". >Of course, some people think that calling bare bsdlabel disks >"dangerously dedicated" or using an MBR is controversial. Actually, you can count me among the folks who think that the adjective "dangerously" may be stretching it a bit, in this context. ___ 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: Advanced Format Drive ?
On Thu, 15 Nov 2012, Ronald F. Guilmette wrote: In message , Warren Block wrote: On Thu, 15 Nov 2012, Ronald F. Guilmette wrote: In your tutorial document, you say: "Create a boot partition to hold the loader, size of 512K." How big is that thing (gpart boot loader), actually? Half a megabyte seems rather a bit large-ish, certainly relative to ye olde MBR loader, which I gather was limited to... what? 32KB (minus a little for the partition table) ? /boot/gptboot is 15K, /boot/gptzfsboot is 39K. A code limitation makes 512K the largest this partition can be made. So I make it that big so it won't have to be increased for bigger boot loaders later. It wouldn't hurt to add the above info to your tutorial page. The problem with that sort of detail is that too much of it obscures the point, which in this case is just trying to show the right way to set up disks without overwhelming the reader. And the space is not wasted because of the next partition... Huh? Oh, oh, OK. I read down to the end of your message and now understand. You meant that the space is not wasted because you will subsequently arrange to have the next following partition begin at 1MB, yes? Yes. Besides which, you probably want to get your readers in the habit of doing things generally on 4KB boundaries, because (as I have just learned) they are probably going to need to start doing that before too long, even if they don't already need to just yet. It does do that, although it's not overt. Starting the first filesystem partition at 1M is a semi-standard, used by various vendors including Microsoft. Yeabut why or how does Microsoft get involved at all with the position of my *FreeBSD* partition?!? There are other vendors and some RAID systems that also use 1M as a starting point. Sticking to that de facto standard helps keep us as compatible as possible with other systems and partitioning software. The cost in space is tiny, and it's a lot easier to do when setting up the disk than after the filesystems are populated. If I began my FreeBSD partition at, say 768KB, would anything from Microsoft be even likely to even notice? I can't say I've tested it. I see this as low-cost insurance. For less than 1M of space, try to be as compatible with other systems and software that exists. Besides being aligned to 4K, it's also aligned to bigger values that can be important for performance on devices like SSDs. I see. That's also another useful tidbit of knowledge that you may also wish to impart to readers of your tutorial. I can only speak for myself, but I for one (perhaps because I have never owned an SSD myself) was totally unaware that those had any such additional alignment issues. Again, I'm trying to avoid too much of that type of detail in that particular article. I've considered writing a separate SSD article, but have not done it yet. P.S. I really do think that it is a serious omission that the gpart(8) man page doesn't really say anything regarding proper or desirable use of the "add -a" option. If it were up to me, I'd put in just a couple of short notes that would say at least something about 4K being Good and Desirable for modern drives, and 1M being Good and Desirable for SSDs. But maybe that kind of info does really belong in something more like an actual tutorial document... you know... something like, um, your's. I can see it both ways. A short mention of those values in that section of gpart(8) would be helpful. The 1M value is controversial to some people. Of course, some people think that calling bare bsdlabel disks "dangerously dedicated" or using an MBR is controversial. ___ 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: Advanced Format Drive ?
In message , Warren Block wrote: >On Thu, 15 Nov 2012, Ronald F. Guilmette wrote: >> Well, given that newfs has been ``fixed'' so that its defaults will >> Do The Right Thing with the latest generation of (4KB block) disks, >> I for one would like to register my vote for fdisk and bsdlabel to >> either (a) be likewise fixed so that they also will default to Doing >> The Right Thing (with the current generation of disks) or else (b) >> be removed from future releases, based on the fact that (apparently) >> they are now so old that nobody cares about them anymore and/or that >> their defaults, when (foolishly?) relied upon, are likely to produce >> Bad Performance, aka Bad Behavior. > >It's legacy code, and that's always a tough call: update and lose the >legacy, or leave it alone and increasingly less useful. Since gpart is >available, there's little pressure to change fdisk or bsdlabel. Well, I'll tell you seriously that I, for one, "didn't get the memo" as the saying goes. Honestly, this discussion is the first time that I personally ever heard that fdisk and/or bsdlabel were being relegated to the dustbin of history. (But then again, I don't get out much, or enough, it seems.) Maybe the man pages should contain notes/warnings saying explicitly "This tool is now depreciated in favor of gpart." What do you think? Is that a suggestion worthy of a formal PR? >> And also, please don't forget the other points I mentioned, i.e. that >> the man page for fdisk makes several references to alignment on ``cylinder'' >> and/or ``head'' boundaries. Are those things even relavant anymore? >> Have they been, anytime in the past 10+ years? (I am guessing that >> there may be other similarly antiquated references to boundaries that >> haven't been meaningful for a long long time also in the bdslabel man >> page, although I confess that didn't even look.) > >I'd say closer to 20 years. Um yea. That's probably closer to the mark. Sigh. Time flies when you're having fun. (And they also like arrows, I'm told.) >But again, it's that legacy thing. And >with FreeBSD, the odds are pretty good that somebody is still running >legacy hardware. Yea, you are undoubtedly right about that. I wonder... can FreeBSD still run on 386s? I can envision a humorous boot-time message that somebody may see someday... Sorry, FreeBSD cannot run on this hardware. Please invest in something that was actually manufactured this century (20xx). :-) ___ 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: Advanced Format Drive ? GPT ?
Warren Block wrote: On Wed, 14 Nov 2012, Ronald F. Guilmette wrote: I'm looking at the examples section of the gpart(8) man page. May I assume that if I just want to merely ``try out'' GPT... you know... taking it out on the road for a first time test run... that I can just do the first five (5) commands listed under EXAMPLES and then that will be enough to go ahead and try installing FreeBSD into the created freebsd-ufs partition? Even assuming that the answer is yes, I have still more questions... Where are these magic numbers coming from?? I am specifically talking about the number "34" in the "-b 34" option and also the number "162" in the "-b 162" option. Tha man page just tosses those into the example command lines without saying a word about them. And you can probably guess what it is that is especially troubling to me about them... neither one of them is divisible by 8 (i.e. 4KB/512B). So would the examples in the current gpart(8) man page produce an Epic Fail when and if they were used with a modern "Advanced Format" drive? -b is the beginning block of a partition. 34 is a magic value, the size of a standard GPT partition table. A good overall reference on GPT is the Wikipedia page: http://en.wikipedia.org/wiki/GUID_Partition_Table Remember that the man page is a reference, not a tutorial. I wanted more specific notes that followed best practices, and that was the source for this article: http://www.wonkity.com/~wblock/docs/html/disksetup.html In general, you create a "partition scheme" first. This can be MBR, GPT, or others. (But use GPT.) Rather than combine the bootcode with the partition table, GPT just uses a small partition for it. Since the standard GPT allows for up to 128 partitions, there's no reason not to use them. Next come other partitions for UFS or ZFS filesystems or swap. That's it, really. The rest is details the man page can explain, like additional options for alignment. (The creation of the first UFS partition in the article does not use -a because older versions of gpart did unexpected things when -a and -b were combined. The alignment produced is correct.) ___ 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" Aloha Warren, I looked over the GPT sample and have a question. In the fstab entries, something that uses msdosfs, (thumb drive maybe). Can you enter it directly in the fstab after the basic partitions and other /dev have been entered in the initial setup? Thanks. ~Al Plant - Honolulu, Hawaii - Phone: 808-284-2740 + http://hawaiidakine.com + http://freebsdinfo.org + + http://aloha50.net - Supporting - FreeBSD 7.2 - 8.0 - 9* + < email: n...@hdk5.net > "All that's really worth doing is what we do for others."- Lewis Carrol ___ 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: Advanced Format Drive ?
In message , Warren Block wrote: >On Thu, 15 Nov 2012, Ronald F. Guilmette wrote: >> In your tutorial document, you say: >> >> "Create a boot partition to hold the loader, size of 512K." >> >> How big is that thing (gpart boot loader), actually? Half a megabyte >> seems rather a bit large-ish, certainly relative to ye olde MBR loader, >> which I gather was limited to... what? 32KB (minus a little for the >> partition table) ? > >/boot/gptboot is 15K, /boot/gptzfsboot is 39K. A code limitation makes >512K the largest this partition can be made. So I make it that big so >it won't have to be increased for bigger boot loaders later. It wouldn't hurt to add the above info to your tutorial page. >And the space is not wasted because of the next partition... Huh? Oh, oh, OK. I read down to the end of your message and now understand. You meant that the space is not wasted because you will subsequently arrange to have the next following partition begin at 1MB, yes? >> Also, when creating the partition to hold the GPT boot loader, shouldn't >> that "gpart add" operation include a "-b 4k" option, you know, on a >> modern "Advanced Format" disk? If not, why not? > >-a 4k, yes. Yea. Sorry. That's what I meant... -a 4k. >It doesn't really matter. The loader is read only at boot, >once, and it's tiny. So it doesn't really matter if it reads at >30M/second or 500M/second. OK. I understand. Thanks. (I did supect that this was the rationale.) >But yes, for consistency, I'll modify that >so the start of the freebsd-boot partition is at 40. It looks prettier that way. Besides which, you probably want to get your readers in the habit of doing things generally on 4KB boundaries, because (as I have just learned) they are probably going to need to start doing that before too long, even if they don't already need to just yet. >> You also go on to say: >> >> "Create partition for /. It should start at the 1M boundary for proper >> sector alignment on 4K sector drives." >> >> Come again? Sorry, but you just lost me entirely. In order to get "proper >> sector alignment" on one of these newer Advanced Format (4k) drives, why >> on earth should it be necessary to begin a partition at some alignment >> which is greater than the obvious minimum, i.e. 4KB ? > >Starting the first filesystem partition at 1M is a semi-standard, used >by various vendors including Microsoft. Yeabut why or how does Microsoft get involved at all with the position of my *FreeBSD* partition?!? If I began my FreeBSD partition at, say 768KB, would anything from Microsoft be even likely to even notice? >Besides being aligned to 4K, it's also aligned to bigger values >that can be important for performance on devices like SSDs. I see. That's also another useful tidbit of knowledge that you may also wish to impart to readers of your tutorial. I can only speak for myself, but I for one (perhaps because I have never owned an SSD myself) was totally unaware that those had any such additional alignment issues. >And that explains the oversized boot partition. It's space that would >be unused otherwise. Got it. Thanks. Regards, rfg P.S. I really do think that it is a serious omission that the gpart(8) man page doesn't really say anything regarding proper or desirable use of the "add -a" option. If it were up to me, I'd put in just a couple of short notes that would say at least something about 4K being Good and Desirable for modern drives, and 1M being Good and Desirable for SSDs. But maybe that kind of info does really belong in something more like an actual tutorial document... you know... something like, um, your's. ___ 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: Advanced Format Drive ?
In message <50a4f2c8.5040...@qeng-ho.org>, Arthur Chance wrote: >On 11/15/12 12:41, Ronald F. Guilmette wrote: >>> -b is the beginning block of a partition. 34 is a magic value, the size >>> of a standard GPT partition table. >> >> It probably wouldn't have hurt anything to mention that in the gpart man >> page. >> >> And what about 162? Is that magic too? If so, how? I seriously do not >> know. > >The man example should be taken as a whole. You've got > >/sbin/gpart add -b 34 -s 128 -t freebsd-boot ad0 > >which gives you a 128 block partition starting at block 34, so the next >free block is 162, and the next partition is explicitly started there in > >/sbin/gpart add -b 162 -s 1048576 -t freebsd-ufs ad0 > >No magic, just arithmetic. :-) Ah! Silly me! Thank you. (I did miss that.) Regards, rfg ___ 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: Advanced Format Drive ?
On Thu, 15 Nov 2012, Ronald F. Guilmette wrote: In message , Warren Block wrote: On Wed, 14 Nov 2012, Ronald F. Guilmette wrote: ... Given these facts, I am more than a little surpised to learn (or rather just to realize) that the good old traditional fdisk and bsdlabel tools do not have ways to explicitly specify minimum alignment _and_ that these tools are still being distributed with FreeBSD. There may be a way, I haven't bothered to look. As I said, gpart does everything fdisk and bsdlabel can do. Well, given that newfs has been ``fixed'' so that its defaults will Do The Right Thing with the latest generation of (4KB block) disks, I for one would like to register my vote for fdisk and bsdlabel to either (a) be likewise fixed so that they also will default to Doing The Right Thing (with the current generation of disks) or else (b) be removed from future releases, based on the fact that (apparently) they are now so old that nobody cares about them anymore and/or that their defaults, when (foolishly?) relied upon, are likely to produce Bad Performance, aka Bad Behavior. It's legacy code, and that's always a tough call: update and lose the legacy, or leave it alone and increasingly less useful. Since gpart is available, there's little pressure to change fdisk or bsdlabel. And also, please don't forget the other points I mentioned, i.e. that the man page for fdisk makes several references to alignment on ``cylinder'' and/or ``head'' boundaries. Are those things even relavant anymore? Have they been, anytime in the past 10+ years? (I am guessing that there may be other similarly antiquated references to boundaries that haven't been meaningful for a long long time also in the bdslabel man page, although I confess that didn't even look.) I'd say closer to 20 years. But again, it's that legacy thing. And with FreeBSD, the odds are pretty good that somebody is still running legacy hardware. ___ 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: Advanced Format Drive ?
On Thu, 15 Nov 2012, Ronald F. Guilmette wrote: In your tutorial document, you say: "Create a boot partition to hold the loader, size of 512K." How big is that thing (gpart boot loader), actually? Half a megabyte seems rather a bit large-ish, certainly relative to ye olde MBR loader, which I gather was limited to... what? 32KB (minus a little for the partition table) ? /boot/gptboot is 15K, /boot/gptzfsboot is 39K. A code limitation makes 512K the largest this partition can be made. So I make it that big so it won't have to be increased for bigger boot loaders later. And the space is not wasted because of the next partition... Also, when creating the partition to hold the GPT boot loader, shouldn't that "gpart add" operation include a "-b 4k" option, you know, on a modern "Advanced Format" disk? If not, why not? -a 4k, yes. It doesn't really matter. The loader is read only at boot, once, and it's tiny. So it doesn't really matter if it reads at 30M/second or 500M/second. But yes, for consistency, I'll modify that so the start of the freebsd-boot partition is at 40. You also go on to say: "Create partition for /. It should start at the 1M boundary for proper sector alignment on 4K sector drives." Come again? Sorry, but you just lost me entirely. In order to get "proper sector alignment" on one of these newer Advanced Format (4k) drives, why on earth should it be necessary to begin a partition at some alignment which is greater than the obvious minimum, i.e. 4KB ? Starting the first filesystem partition at 1M is a semi-standard, used by various vendors including Microsoft. Besides being aligned to 4K, it's also aligned to bigger values that can be important for performance on devices like SSDs. And that explains the oversized boot partition. It's space that would be unused otherwise. ___ 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: Advanced Format Drive ?
On Thu, 15 Nov 2012, Chris Whitehouse wrote: In message, In general, you create a "partition scheme" first. This can be MBR, GPT, or others. (But use GPT.) Unless you want to dual boot with WinXP in which case use MBR still? Yes. The same for Vista or Windows 7, mostly. AFAIK, Windows 7 64-bit on a UEFI system is the only Windows that will boot from GPT. As I've said before, consider using VMs rather than dual booting. ___ 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: Advanced Format Drive ?
On 11/15/12 12:41, Ronald F. Guilmette wrote: In message , Warren Block wrote: On Wed, 14 Nov 2012, Ronald F. Guilmette wrote: I'm looking at the examples section of the gpart(8) man page. May I assume that if I just want to merely ``try out'' GPT... you know... taking it out on the road for a first time test run... that I can just do the first five (5) commands listed under EXAMPLES and then that will be enough to go ahead and try installing FreeBSD into the created freebsd-ufs partition? Even assuming that the answer is yes, I have still more questions... Where are these magic numbers coming from?? I am specifically talking about the number "34" in the "-b 34" option and also the number "162" in the "-b 162" option. Tha man page just tosses those into the example command lines without saying a word about them. And you can probably guess what it is that is especially troubling to me about them... neither one of them is divisible by 8 (i.e. 4KB/512B). So would the examples in the current gpart(8) man page produce an Epic Fail when and if they were used with a modern "Advanced Format" drive? -b is the beginning block of a partition. 34 is a magic value, the size of a standard GPT partition table. It probably wouldn't have hurt anything to mention that in the gpart man page. And what about 162? Is that magic too? If so, how? I seriously do not know. The man example should be taken as a whole. You've got /sbin/gpart add -b 34 -s 128 -t freebsd-boot ad0 which gives you a 128 block partition starting at block 34, so the next free block is 162, and the next partition is explicitly started there in /sbin/gpart add -b 162 -s 1048576 -t freebsd-ufs ad0 No magic, just arithmetic. :-) ___ 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: Advanced Format Drive ?
In message, In general, you create a "partition scheme" first. This can be MBR, GPT, or others. (But use GPT.) Unless you want to dual boot with WinXP in which case use MBR still? Chris ___ 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: Advanced Format Drive ?
In message , Warren Block wrote: >On Wed, 14 Nov 2012, Ronald F. Guilmette wrote: > >> I'm looking at the examples section of the gpart(8) man page. May I >> assume that if I just want to merely ``try out'' GPT... you know... >> taking it out on the road for a first time test run... that I can >> just do the first five (5) commands listed under EXAMPLES and then >> that will be enough to go ahead and try installing FreeBSD into the >> created freebsd-ufs partition? >> >> Even assuming that the answer is yes, I have still more questions... >> Where are these magic numbers coming from?? I am specifically talking >> about the number "34" in the "-b 34" option and also the number "162" >> in the "-b 162" option. Tha man page just tosses those into the example >> command lines without saying a word about them. And you can probably >> guess what it is that is especially troubling to me about them... neither >> one of them is divisible by 8 (i.e. 4KB/512B). So would the examples >> in the current gpart(8) man page produce an Epic Fail when and if they >> were used with a modern "Advanced Format" drive? > >-b is the beginning block of a partition. 34 is a magic value, the size >of a standard GPT partition table. It probably wouldn't have hurt anything to mention that in the gpart man page. And what about 162? Is that magic too? If so, how? I seriously do not know. >A good overall reference on GPT is >the Wikipedia page: http://en.wikipedia.org/wiki/GUID_Partition_Table > >Remember that the man page is a reference, not a tutorial. Actually, it is clearly both. We all know that man pages are primarily supposed to be (minimal?) reference documents, but you cannot claim with a straight face that any man page that contains an EXAMPLES section is not also serving as a rudimentary tutorial. Personally, I find the minimalist tutorials that are often found within EXAMPLES sections of man page quite helpful, gpart(8) included. But in this specific case the pulling of number, apparently out of thin air (at least from the point of view of the uninitiated) rather significantly degrades the educational value. >I wanted >more specific notes that followed best practices, and that was the >source for this article: >http://www.wonkity.com/~wblock/docs/html/disksetup.html It is very helpful (and very appreciated!) that you were kind enough to create that document, which is clearly more unambiguously a tutorial. But really, the gpart(8) man page got me about 97% of the way there, even without me having to consult external references. If it just had not been for those mystery numbers... >In general, you create a "partition scheme" first. This can be MBR, >GPT, or others. (But use GPT.) Yea. I got that part. >Rather than combine the bootcode with the partition table, GPT just uses >a small partition for it. Since the standard GPT allows for up to 128 >partitions, there's no reason not to use them. Got it. Thanks. >Next come other partitions for UFS or ZFS filesystems or swap. Right. >That's it, really. The rest is details the man page can explain, like >additional options for alignment. (The creation of the first UFS >partition in the article does not use -a because older versions of gpart >did unexpected things when -a and -b were combined. The alignment >produced is correct.) Questions: In your tutorial document, you say: "Create a boot partition to hold the loader, size of 512K." How big is that thing (gpart boot loader), actually? Half a megabyte seems rather a bit large-ish, certainly relative to ye olde MBR loader, which I gather was limited to... what? 32KB (minus a little for the partition table) ? Also, when creating the partition to hold the GPT boot loader, shouldn't that "gpart add" operation include a "-b 4k" option, you know, on a modern "Advanced Format" disk? If not, why not? You also go on to say: "Create partition for /. It should start at the 1M boundary for proper sector alignment on 4K sector drives." Come again? Sorry, but you just lost me entirely. In order to get "proper sector alignment" on one of these newer Advanced Format (4k) drives, why on earth should it be necessary to begin a partition at some alignment which is greater than the obvious minimum, i.e. 4KB ? ___ 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: Advanced Format Drive ?
In message , Warren Block wrote: >On Wed, 14 Nov 2012, Ronald F. Guilmette wrote: >... >> Given these facts, I am more than a little surpised to learn (or rather >> just to realize) that the good old traditional fdisk and bsdlabel tools >> do not have ways to explicitly specify minimum alignment _and_ that >> these tools are still being distributed with FreeBSD. > >There may be a way, I haven't bothered to look. As I said, gpart does >everything fdisk and bsdlabel can do. Well, given that newfs has been ``fixed'' so that its defaults will Do The Right Thing with the latest generation of (4KB block) disks, I for one would like to register my vote for fdisk and bsdlabel to either (a) be likewise fixed so that they also will default to Doing The Right Thing (with the current generation of disks) or else (b) be removed from future releases, based on the fact that (apparently) they are now so old that nobody cares about them anymore and/or that their defaults, when (foolishly?) relied upon, are likely to produce Bad Performance, aka Bad Behavior. And also, please don't forget the other points I mentioned, i.e. that the man page for fdisk makes several references to alignment on ``cylinder'' and/or ``head'' boundaries. Are those things even relavant anymore? Have they been, anytime in the past 10+ years? (I am guessing that there may be other similarly antiquated references to boundaries that haven't been meaningful for a long long time also in the bdslabel man page, although I confess that didn't even look.) ___ 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: Advanced Format Drive ?
On Wed, 14 Nov 2012, Ronald F. Guilmette wrote: I'm looking at the examples section of the gpart(8) man page. May I assume that if I just want to merely ``try out'' GPT... you know... taking it out on the road for a first time test run... that I can just do the first five (5) commands listed under EXAMPLES and then that will be enough to go ahead and try installing FreeBSD into the created freebsd-ufs partition? Even assuming that the answer is yes, I have still more questions... Where are these magic numbers coming from?? I am specifically talking about the number "34" in the "-b 34" option and also the number "162" in the "-b 162" option. Tha man page just tosses those into the example command lines without saying a word about them. And you can probably guess what it is that is especially troubling to me about them... neither one of them is divisible by 8 (i.e. 4KB/512B). So would the examples in the current gpart(8) man page produce an Epic Fail when and if they were used with a modern "Advanced Format" drive? -b is the beginning block of a partition. 34 is a magic value, the size of a standard GPT partition table. A good overall reference on GPT is the Wikipedia page: http://en.wikipedia.org/wiki/GUID_Partition_Table Remember that the man page is a reference, not a tutorial. I wanted more specific notes that followed best practices, and that was the source for this article: http://www.wonkity.com/~wblock/docs/html/disksetup.html In general, you create a "partition scheme" first. This can be MBR, GPT, or others. (But use GPT.) Rather than combine the bootcode with the partition table, GPT just uses a small partition for it. Since the standard GPT allows for up to 128 partitions, there's no reason not to use them. Next come other partitions for UFS or ZFS filesystems or swap. That's it, really. The rest is details the man page can explain, like additional options for alignment. (The creation of the first UFS partition in the article does not use -a because older versions of gpart did unexpected things when -a and -b were combined. The alignment produced is correct.) ___ 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: Advanced Format Drive ?
On Wed, 14 Nov 2012, Ronald F. Guilmette wrote: (And I gather from everything that has been said so far in this thread that if the alignment is set wrong, then the user is likely to pay a Big Price in terms of performance, right?) Yes. ... and I am almost tempted to file a formal PR about this, i.e. the fact that ``guided'' partitioning doesn't allow the user to specify the alignment of _anything_. There are a couple PRs like that already. Or do I need to set the alignment separately, e.g. my manually running bsdlabel? (Normally, I've just been using what noadays is being called "guided" partitioning, you know, with the friendly curses-based GUI. So As with fdisk, I have no real experience using bsdlabee from teh command line. But I guess it is time that i learned how.) I don't know of a way to make fdisk and bsdlabel do the correct alignment. That also is rather entirely perplexing to me, especially given all else that I have learned already from and within this conversation. fdisk and bsdlabel are old tools. Disks have had 512-byte blocks for a very long time. For example, I've learned that when one is using modern "Advanced Format) (4KB blocksize) hard disks, it is Bad (capital `B') to allow any partition to be aligned to anything other than (at least) a 4KB boundary, _and_ that newfs has already, apparently been modified/updated so that it's minimum default fragment size is 4KB. The larger size was an option to newfs, the defaults have just been changed. Given these facts, I am more than a little surpised to learn (or rather just to realize) that the good old traditional fdisk and bsdlabel tools do not have ways to explicitly specify minimum alignment _and_ that these tools are still being distributed with FreeBSD. There may be a way, I haven't bothered to look. As I said, gpart does everything fdisk and bsdlabel can do. ___ 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: Advanced Format Drive ?
geezer, and I _do_ remember the ancient times when disk drives actually told the real unvarnished truth, and when OSes actually did go to some lengths to try to implement optimized disk usage strategies, e.g. to try to keep the heads as close to the centers of their ranges as possible, and so forth. But I gather that all this stuff went the way of the dinosaurs long long ago, and thus, my questions.) >But that's okay, because gpart(8) does everything they do, >and more. Creating MBRs and bsdlabels is more work, but gpart can do >it, and do the juggling to get the bsdlabel partitions to line up. Reading the gpart(8) man page, I see that you are (of course) 100% correct. It can do all I might need and more. But I also have a number of questions about gpart. (See below.) >Again, I suggest that GPT is the much easier and more versatile way. OK, I'm conservative in my approach to technology generally, and I've learned the hard way not to try most new ``innovations'' until at least v2.0, by which time, hopefully, all of the kinks have been worked out. But GPT has been around for awhile now and I'm willing to take a crack at this new way of partitioning. And I've already put a fresh blank disk in one of my machines that I'm ready to try it on, but... I'm looking at the examples section of the gpart(8) man page. May I assume that if I just want to merely ``try out'' GPT... you know... taking it out on the road for a first time test run... that I can just do the first five (5) commands listed under EXAMPLES and then that will be enough to go ahead and try installing FreeBSD into the created freebsd-ufs partition? Even assuming that the answer is yes, I have still more questions... Where are these magic numbers coming from?? I am specifically talking about the number "34" in the "-b 34" option and also the number "162" in the "-b 162" option. Tha man page just tosses those into the example command lines without saying a word about them. And you can probably guess what it is that is especially troubling to me about them... neither one of them is divisible by 8 (i.e. 4KB/512B). So would the examples in the current gpart(8) man page produce an Epic Fail when and if they were used with a modern "Advanced Format" drive? >But if you insist on MBR/bsdlabel... No, actually, now I am motivated to make the leap to using GPT... *if* I can just manage to make heads or tails of it, that is. But this whole alignment issue is now making me go back (e.g. to the older tools fdisk and bsdlabel) and question whether I even ever properly understood what _that stuff_ was doing (and more importantly *why* it was doing it) with respect to alignment. Regards, rfg ___ 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: Advanced Format Drive ?
On Tue, 13 Nov 2012, Ronald F. Guilmette wrote: OK. I think that I always was doing that anyway. But I want to be sure that I understand... If the size of the BSD partition is a multiple of, say, !MB, then the _alignment_ of that partition will likewise (auto- magically) be at least 1MB also? No. If you start with $0.63, and only add full dollars or tens, you will still never have an integer amount of dollars. Or do I need to set the alignment separately, e.g. my manually running bsdlabel? (Normally, I've just been using what noadays is being called "guided" partitioning, you know, with the friendly curses-based GUI. So As with fdisk, I have no real experience using bsdlabee from teh command line. But I guess it is time that i learned how.) I don't know of a way to make fdisk and bsdlabel do the correct alignment. But that's okay, because gpart(8) does everything they do, and more. Creating MBRs and bsdlabels is more work, but gpart can do it, and do the juggling to get the bsdlabel partitions to line up. Again, I suggest that GPT is the much easier and more versatile way. But if you insist on MBR/bsdlabel, there are examples of creating it with gpart in the new gmirror section of the GEOM chapter of the Handbook: http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/geom-mirror.html ___ 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: Advanced Format Drive ?
On Tue, Nov 13, 2012 at 2:07 PM, Ronald F. Guilmette wrote: > >You can then easily use newfs with the -f parameter: > > > > newfs -U -f 4096 > > > >This will make sure the proper fragment size will be applied > >upon formatting the created partitions. > > OK. Thanks. I am guessing that this is really the one and probably > _only_ thing that might really make any significant difference, > performance- > wise, right? I mean if the partition is improperly aligned, that really > only would affect reading and/or writing at the very beginning or at the > very end of the partition, right? No, I think it'll affect the whole thing, because every block will be straddling two 4k sectors instead of filling one. Then the drive has to read and rewrite two sectors for every one block you write, instead of just one, slowing things down to half speed. Imagine running across the top of a picket fence...if you start half a fencepost off things are going to be painful. ;) ___ 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: Advanced Format Drive ?
> From: "Ronald F. Guilmette" > Subject: Re: Advanced Format Drive ? > Date: Tue, 13 Nov 2012 14:07:50 -0800 > > > And while we are on the subject... Has anybody ever down any analysis > (i.e. benchmarking) to find out if -f 4096 is even the best number for > a modern high(er) capacity drive? I'm just sort-of wondering if 8192 > or 16384 might be better. As long as the fragment size is a power-of-two multiple of the media sector size, there is no significant performance difference. The only case where a larger fragment size makes any difference is heavy random i/o on files where the larger fragment size translates to one less level of indirect block in the meta-data. Larger fragment sizes also make for more 'waste' space in the 'used' part of the disk, assuming random file sizes. And reduce the space savings gained by _not_ writing 'holes' to disk. For these reasons, in general, small is better. > 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: Advanced Format Drive ?
In message <50a2002b.9040...@qeng-ho.org>, Arthur Chance wrote: >According to the manual as of 9.0-RELEASE the default fragment and block >sizes for newfs are 4k and 32k, so provided your partitions/slices are >4k aligned everything Should Just Work. Before 9.0 fragments and blocks >were 2k and 16k which doesn't play so well with 4k drives. Thank you Arthur for pointing this out. This comes as welcome news, since now, it would seem, I won't have to get down a grunge around trying to run the command line versions of fdisk, bsdlabel, and newfs. (Normally, I prefer doing most things from the command line, but initializing new disks for use with FreeBSD is one of the rare exceptional cases where I prefer to have a bit of a GUI wrapper that's double checking to make sure that I don't do anything completely goofy.) ___ 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: Advanced Format Drive ?
In message <20121113073030.87bc0608.free...@edvax.de>, Polytropon wrote: >Note that 4k = 8 x 512 byte, and so 64 sectors would be a >good alignment "grid", while 63 sectors is not. That implies >that in case you use fdisk to create a slice holding your >partitions, try to make it start at sector 64 (63 would >have been the default). OK. I've only ever used the FreeBSD fdisk to just look at what the current (DOS) partitioning is, so I guess I'll have to dig into the man page and try to figure out how to actually use it to create a DOS partition starting at "block" 64. >After that, use bsdlabel to create the partitions inside >the slice as you want. Make them multiples of 1M or 1G, OK. I think that I always was doing that anyway. But I want to be sure that I understand... If the size of the BSD partition is a multiple of, say, !MB, then the _alignment_ of that partition will likewise (auto- magically) be at least 1MB also? Or do I need to set the alignment separately, e.g. my manually running bsdlabel? (Normally, I've just been using what noadays is being called "guided" partitioning, you know, with the friendly curses-based GUI. So As with fdisk, I have no real experience using bsdlabee from teh command line. But I guess it is time that i learned how.) >that should be no big deal because disks are big and cheap >today. :-) Yes, exactly so. I am not exactly going to sweat losing even, say, one megabyte now that I am the proud owner of a shiny new one TERABYTE drive. (Thirty years ago, I could hardly have even ever imagined that such might exist one day, let alone that I myself would own one, and let alone that I might have been able to purchase one for less than $100 USD. Rather amazing really.) >You can then easily use newfs with the -f parameter: > > newfs -U -f 4096 > >This will make sure the proper fragment size will be applied >upon formatting the created partitions. OK. Thanks. I am guessing that this is really the one and probably _only_ thing that might really make any significant difference, performance- wise, right? I mean if the partition is improperly aligned, that really only would affect reading and/or writing at the very beginning or at the very end of the partition, right? Whereas this -f parameter for newfs is, I gather, the thing that really tells the kernel the size of the physical chunks of data that it can/should read/write to the drive at any one time, right? And while we are on the subject... Has anybody ever down any analysis (i.e. benchmarking) to find out if -f 4096 is even the best number for a modern high(er) capacity drive? I'm just sort-of wondering if 8192 or 16384 might be better. ___ 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: Advanced Format Drive ?
On Mon, 12 Nov 2012, Ronald F. Guilmette wrote: In message <20121113065602.ee2310d7.free...@edvax.de>, Polytropon wrote: On Mon, 12 Nov 2012 21:47:40 -0800, Ronald F. Guilmette wrote: Is there _anything_ that I will have to do differently than I did for the last 20 drives I've used with FreeBSD over the last 10+ years? As far as I know, the "old ways" still work as intended. I've been initalizing 1 TB and 1.5 TB disks the "old way", using sysinstall (to create a slice, then to create the partitions) and newfs (to format the 2nd data disk). So far, the disks are working for some years without trouble. Those are "normal" disks, not SSDs, purchased few years ago. The term "advanced format" is usually used for 4k-sectorized disks (in difference to "traditional" 512k sectors). You can find more here: http://en.wikipedia.org/wiki/Advanced_Format The implication for FreeBSD is (and has been for some time) to align partitions "at a 4k border". If you create partition sizes as multiples of 4k, it should be fine. Thank you. Which "partitions" need to be aligned to the 4KB boundaries? The FreeBSD ones, the MBR ones, or both? The ones you want to be fast instead of half-speed. The easy way to do it is to jettison the old MBR slice/partition stuff and use the simpler GPT. The first half of this document shows how to set it up: http://www.wonkity.com/~wblock/docs/html/disksetup.html ___ 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: Advanced Format Drive ?
On 11/13/12 06:30, Polytropon wrote: On Mon, 12 Nov 2012 22:14:11 -0800, Ronald F. Guilmette wrote: Which "partitions" need to be aligned to the 4KB boundaries? The FreeBSD ones, the MBR ones, or both? The partitions, all of them. :-) For MBR partitions, the "DOS primary partitions", which are slices, you typically only need one if you want to stay in compatibility mode. For dedicated mode, you don't need it. The slice typically starts in sector 63 and occupies the space until the end of the device. The partitions within the slice should have sizes of multiples of 1 MB or 1 GB (which makes them multiples of 4k easily). See: http://www.freebsd.org/doc/handbook/bsdinstall-partitioning.html Even though the handbook elaborates on the GPT approach, it will work with traditional disklabel partitioning too. Note that 4k = 8 x 512 byte, and so 64 sectors would be a good alignment "grid", while 63 sectors is not. That implies that in case you use fdisk to create a slice holding your partitions, try to make it start at sector 64 (63 would have been the default). After that, use bsdlabel to create the partitions inside the slice as you want. Make them multiples of 1M or 1G, that should be no big deal because disks are big and cheap today. :-) You can then easily use newfs with the -f parameter: newfs -U -f 4096 This will make sure the proper fragment size will be applied upon formatting the created partitions. Also see: http://www.wonkity.com/~wblock/docs/html/disksetup.html As I have limited experience, anyone having more practical experience with this matter is welcome to comment. :-) According to the manual as of 9.0-RELEASE the default fragment and block sizes for newfs are 4k and 32k, so provided your partitions/slices are 4k aligned everything Should Just Work. Before 9.0 fragments and blocks were 2k and 16k which doesn't play so well with 4k drives. ___ 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: Advanced Format Drive ?
On Mon, 12 Nov 2012 22:14:11 -0800, Ronald F. Guilmette wrote: > Which "partitions" need to be aligned to the 4KB boundaries? > The FreeBSD ones, the MBR ones, or both? The partitions, all of them. :-) For MBR partitions, the "DOS primary partitions", which are slices, you typically only need one if you want to stay in compatibility mode. For dedicated mode, you don't need it. The slice typically starts in sector 63 and occupies the space until the end of the device. The partitions within the slice should have sizes of multiples of 1 MB or 1 GB (which makes them multiples of 4k easily). See: http://www.freebsd.org/doc/handbook/bsdinstall-partitioning.html Even though the handbook elaborates on the GPT approach, it will work with traditional disklabel partitioning too. Note that 4k = 8 x 512 byte, and so 64 sectors would be a good alignment "grid", while 63 sectors is not. That implies that in case you use fdisk to create a slice holding your partitions, try to make it start at sector 64 (63 would have been the default). After that, use bsdlabel to create the partitions inside the slice as you want. Make them multiples of 1M or 1G, that should be no big deal because disks are big and cheap today. :-) You can then easily use newfs with the -f parameter: newfs -U -f 4096 This will make sure the proper fragment size will be applied upon formatting the created partitions. Also see: http://www.wonkity.com/~wblock/docs/html/disksetup.html As I have limited experience, anyone having more practical experience with this matter is welcome to comment. :-) -- 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: Advanced Format Drive ?
In message <20121113065602.ee2310d7.free...@edvax.de>, Polytropon wrote: >On Mon, 12 Nov 2012 21:47:40 -0800, Ronald F. Guilmette wrote: >> Is there _anything_ that I will have to do differently than I did for the >> last 20 drives I've used with FreeBSD over the last 10+ years? > >As far as I know, the "old ways" still work as intended. >I've been initalizing 1 TB and 1.5 TB disks the "old way", >using sysinstall (to create a slice, then to create the >partitions) and newfs (to format the 2nd data disk). So >far, the disks are working for some years without trouble. >Those are "normal" disks, not SSDs, purchased few years >ago. > >The term "advanced format" is usually used for 4k-sectorized >disks (in difference to "traditional" 512k sectors). > >You can find more here: >http://en.wikipedia.org/wiki/Advanced_Format > >The implication for FreeBSD is (and has been for some time) >to align partitions "at a 4k border". If you create partition >sizes as multiples of 4k, it should be fine. Thank you. Which "partitions" need to be aligned to the 4KB boundaries? The FreeBSD ones, the MBR ones, or both? ___ 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: Advanced Format Drive ?
On Mon, 12 Nov 2012 21:47:40 -0800, Ronald F. Guilmette wrote: > Is there _anything_ that I will have to do differently than I did for the > last 20 drives I've used with FreeBSD over the last 10+ years? As far as I know, the "old ways" still work as intended. I've been initalizing 1 TB and 1.5 TB disks the "old way", using sysinstall (to create a slice, then to create the partitions) and newfs (to format the 2nd data disk). So far, the disks are working for some years without trouble. Those are "normal" disks, not SSDs, purchased few years ago. The term "advanced format" is usually used for 4k-sectorized disks (in difference to "traditional" 512k sectors). You can find more here: http://en.wikipedia.org/wiki/Advanced_Format The implication for FreeBSD is (and has been for some time) to align partitions "at a 4k border". If you create partition sizes as multiples of 4k, it should be fine. -- 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"
Advanced Format Drive ?
OK, so I'm lazy. Guilty as charged. I started to do some googling around to try and find out what the meaning of this warning on the anti-static bag outside of my brand new 1TB hard drive might be, and what the implications might be for FreeBSD, but so far all I am finding is seemingly endless discussions on some of the mailing lists that seem to go on forever without ever reaching any definitive conclusion(s). So could someobody just give it to me straight and briefly? What's the bottom line here? What, if anything, do I have to do _differently_ than what I am accustomed to in order to put a good old fashioned MBR partition table on this thing, and then to create a slice #1 and then use bsdlabel to slice that up in turn into a set of UFS parititions? Is there _anything_ that I will have to do differently than I did for the last 20 drives I've used with FreeBSD over the last 10+ years? Note that I have _no_ intention to use ZFS. (I gather that there may perhaps be some issue with using ZFS together with "Advanced Format" drives.) ___ 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"