Re: LordSutch netinstall CD
Richard Atterer <[EMAIL PROTECTED]> writes: > On Mon, Dec 06, 2004 at 10:34:16PM +0100, Yannick B wrote: > > I am hereby reporting a broken link to LordSutch.com netinstall CD. > > Does anyone know of a mirror of these images? ISTR they worked quite well. > :-/ I'll have to remove the link otherwise. http://ftp.fi.debian.org/ has a mirror of those images. -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: 3.0r3 CDs ready
Steve McIntyre <[EMAIL PROTECTED]> writes: > ISOs and jigdos are now available on cdimage.debian.org and (shortly) > the mirrors for all architectures, along with 3.0->3.0r3 update > CDs. I'm working on DVDs for all architectures now. ftp.fi.debian.org has 3.0r3 CD images available now. The CD image push trigger and the resulting sync worked flawlessly and the sync was finished on Fri Oct 29 06:21:31 UTC 2004. -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: state of the cdrom mirrors report
Joey Hess <[EMAIL PROTECTED]> writes: > A few places have question marks; this is where the file exists, but > seems to have the wrong size. I think some/all of those are really ok, > and are due to weirdness in the ftp SIZE command on some servers. It may be that those servers are functioning correctly. Does you client do something like $ftp->type("I"); to ask for non-translated files? ftp.fi.debian.org was one of the mirrors reporting weird size, so I did some testing. Here is a transcript: ftp> ls 229 Entering Extended Passive Mode (|||60002|) 150 Opening ASCII mode data connection for '/bin/ls'. total 4907684 -rwxr-xr-x 1 deb deb797 Jul 20 19:41 MD5SUMS -rwxr-xr-x 1 deb deb 612171776 Jul 20 19:31 debian-30r0-i386-binary-1.iso ftp> size debian-30r0-i386-binary-1.iso debian-30r0-i386-binary-1.iso 614780752 ftp> bin 200 Type set to I. ftp> size debian-30r0-i386-binary-1.iso debian-30r0-i386-binary-1.iso 612171776 When the type was set to binary, the response came immediately and was correct. Before that, the command took a long time to complete, and I noticed that the ftpd was taking all the CPU time it could get. I did not check the sources, but I bet it was doing the line end translation for ascii mode. -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: A new Finnish ISO image mirror
Richard Atterer <[EMAIL PROTECTED]> writes: > Hm, can anybody confirm that the isos are actually there? The site > really refuses all connections, not just attempts to download the iso > files... ftp://ftp.bittivuoto.net/iso/debian/ seems to contain only the 8 i386 images. Accessing the mirror worked fine from tut.fi. -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: About jigdo test on cdimage.d.o
Philip Hands <[EMAIL PROTECTED]> writes: > On Thu, 2002-05-02 at 18:19, Heikki Vatiainen wrote: > > The images were built correctly once maxMissing was raised above 25, > > just as Mattias noted. It looks like about 80 files were fetched from > > the snapshot archive for e.g. i386 disk 3. This was the reason why I > > only built three CDs since downloading hundreds of files for just > > testing seemed like waste of bandwidth when an up-to-date mirror is > > sitting on the same disk. Phil, are you planning to update the > > snapshot soon? > > Yeah, will do tonight probably. open's currently fscking it's /home, > having marked /home read-only after finding a broken directory inode > this afternoon. Looks like 3.0-pre3 for i386 and source have been made available. I gave them a try and now everything was taken from the local mirror except of one file. This is pool/main/g/geomview/geomview_1.8.1.orig.tar.gz and it looks like you still have a corrupted file in your mirror. Please see my mail to debian-cd on 11 April, you may have missed it then. I am also interested in the general opinion if the test images should be made publicly available. I have only built the images but I have not otherwise tested them by burning them on a CD and trying to do the actual installation. -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
About jigdo test on cdimage.d.o
I built three i386 images with jigdo-mirror using jigdo and template files from http://cdimage.debian.org/jigdo-area/ The images were built correctly once maxMissing was raised above 25, just as Mattias noted. It looks like about 80 files were fetched from the snapshot archive for e.g. i386 disk 3. This was the reason why I only built three CDs since downloading hundreds of files for just testing seemed like waste of bandwidth when an up-to-date mirror is sitting on the same disk. Phil, are you planning to update the snapshot soon? Also, the jigdo and template files seem not to be available via rsync anymore. Will this be changed? -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: isolinux image updated
Chris Lawrence <[EMAIL PROTECTED]> writes: > I've updated the ISOLINUX test image as follows: > - Updated to woody as of Wed, 17 Apr 2002 19:57:19 UTC Would it be possible to add wireless-tools to the package list? It is the only package that would be needed to finish the installation over WLAN. -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: pre-release Jigdo files appearing at cdimage.d.o, please test
Richard Atterer <[EMAIL PROTECTED]> writes: > On Thu, Apr 11, 2002 at 11:37:56AM +0300, Heikki Vatiainen wrote: > > Jigdo is very good in finding out files that differ between > > archives. While generating the 6th CD for i386 I found out I had a > > corrupted file in my local archive. It would be helpful if > > jigdo-lite told what files are missing or had a command lite option > > for doing that. > > Outputting that list is pretty easy: After jigdo-lite tells you that > the download of some files has failed, interrupt it and use > "jigdo-file print-missing -i imagename", it will pick up imagename.tmp > and output the missing files. Thanks, that did the trick. I did experiments with print-missing, but I did not specify imagename but imagename.tmp instead. That only caused jigdo-file to return without any output. -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: pre-release Jigdo files appearing at cdimage.d.o, please test
Philip Hands <[EMAIL PROTECTED]> writes: > I'm just running off some images, which are appearing here: > > http://cdimage.debian.org/pre-release-jigdo/ I pulled the files from rsync::debian-cd-test and used the .jigdo and .template files locally. The i386 build was now successful, you seem to have tweaked the files, and all but the 4th source CD were also built successfully. The problem with the 4th CD seems to be caused by a corrupted file. Here are some notes: The .jigdo files for source CDs are trying to use http://cdimage.debian.org/cd-images/jigdo/templates/3.0BETA/source/ It looks like the last component should be src, not source. Jigdo is very good in finding out files that differ between archives. While generating the 6th CD for i386 I found out I had a corrupted file in my local archive. It would be helpful if jigdo-lite told what files are missing or had a command lite option for doing that. It also looks that you have a corrupted source package in your archive. The 4th source CD failed to build and this is what I found out: % zgrep pool/main/g/geomview/geomview_1.8.1.orig.tar.gz debian/indices/md5sums.gz 1cb0e07a6552e9fe122491af70171536 pool/main/g/geomview/geomview_1.8.1.orig.tar.gz % md5sum pool/main/g/geomview/geomview_1.8.1.orig.tar.gz 1cb0e07a6552e9fe122491af70171536 pool/main/g/geomview/geomview_1.8.1.orig.tar.gz % jigdo-file md5 pool/main/g/geomview/geomview_1.8.1.orig.tar.gz HLDgemVS6f4SJJGvcBcVNg pool/main/g/geomview/geomview_1.8.1.orig.tar.gz % grep pool/main/g/geomview/geomview_1.8.1.orig.tar.gz src/woody-src-4.jigdo ZqEQbYYSbI5w_bce9C4dbw=Debian:pool/main/g/geomview/geomview_1.8.1.orig.tar.gz -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: pre-release Jigdo files appearing at cdimage.d.o, please test
I tried generating the images with jigdo-lite, but run into problems. jigdo-lite was reporting one missing file ("Found 982 of the 983 files required by the template") and was not able to complete the image. This was my first time with jigdo but I suspect the problem was that non-US was updated after you had made the jigdo files. >From woody-i386-1.jigdo: % grep non-US woody-i386-1.jigdo sljMO1JHdJDT_gQYa2IPtg=Non-US:dists/woody/non-US/Contents-i386.gz >From the actual file system: % jigdo-file md5sum dists/woody/non-US/Contents-i386.gz Iegxgg-6cQCWTLwa4XFOrw dists/woody/non-US/Contents-i386.gz Choosing some other file, such as tools/unz512x3.exe, gives the same checksum that is in the .jigdo file. -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
isolinux test image mirror
We also have a mirror of Chris Lawrence's isolinux test image. We currently have the fifth update and intent to poll for new updates. rsync://trumpetti.atm.tut.fi/debian-woody-multiboot http://trumpetti.atm.tut.fi/ftp/debian-woody-multiboot/ ftp://trumpetti.atm.tut.fi/pub/debian-woody-multiboot/ -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Multiboot testing please ...
Chris Lawrence <[EMAIL PROTECTED]> writes: > http://relativity.phy.olemiss.edu/~cnlawren/woody-i386-1.raw Could you give us a MD5SUM for it? I am pulling it from Mattias and thought about mentioning it in our national Usenet linux group. -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: CD image mirrors
Josip Rodin <[EMAIL PROTECTED]> writes: > Here's the data I promised: the CD image mirrors recorded in the central > Debian mirror database. If anything is wrong or missing, please fill us in. > CC:ing corrections to [EMAIL PROTECTED] is desirable... > Site: ftp.fi.debian.org > CDImage-http: /debian-cd/ > CDImage-ftp: /debian-cd/ > CDImage-rsync: ftp.fi.debian.org also has rsync. Please add CDImage-rsync: debian-cd/ Thanks, -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: need cd mirror
Othmar Pasteka <[EMAIL PROTECTED]> wrote: > well, since phil hands' cdimage mirror is too fast (would be too > expensive) i am looking for one who also mirrors the cdimages and > which ideally _does_not_ route through carrier1 to me. i.e. > abovenet and ebone/gts would be great. Try tracerouting to ftp.fi.debian.org (trumpetti.atm.tut.fi). I just got a complete 2.2_rev2 mirror online. At least the trace from Tampere, Finland goes to www.kabsi.at via teleglobe in New York and then through above.net. You can access the mirror with: rsync://ftp.fi.debian.org/debian-cd/ ftp://ftp.fi.debian.org/debian-cd/ http://ftp.fi.debian.org/debian-cd/ -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: where to rsync from
Othmar Pasteka <[EMAIL PROTECTED]> wrote: > who is offering the cdimages via rsync because cdimage.debian.org > is so slow that i will wait weeks till everything is mirrored. Actually cdimage.debian.org seems to work very well now. Even the slow network connection problems are gone. You might want to try trumpetti.atm.tut.fi::debian-cd which has a full mirror. > PS: ;) // Heikki -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Two more corrupt .debs found from 2.2_rev0 images
jason andrade <[EMAIL PROTECTED]> wrote: > out of curiousity - where are the "2.2_rev0_CDa" cds being produced ? I can not give authorative answer since I'm not Phil, but my guess is that they are/were also produced on cdimage.debian.org. > i can't see any on cdimage which i thought was the authoritative > debian cd creation site. are debian-cd mirrors meant to have this > new revision already ? because i haven't updated ours yet and so far > i appear to be the only site in Australia carrying 2.2 images. Maybe you should take another look ;-) c.d.o does have 2.2_rev0_CDa and that's where I got them yesterday. Since they were available, I thought it was appropriate to get them. % date --utc Sat Aug 19 14:16:59 UTC 2000 % rsync -av [EMAIL PROTECTED]::debian-cd |grep CDa drwxr-xr-x4096 2000/08/18 05:52:37 2.2_rev0_CDa drwxr-xr-x4096 2000/08/18 05:51:51 2.2_rev0_CDa/sparc -rw-r--r-- 493 2000/08/18 05:51:30 2.2_rev0_CDa/sparc/MD5SUMS -rw-r--r-- 667688960 2000/08/18 05:26:00 2.2_rev0_CDa/sparc/binary-sparc-1.iso -rw-r--r-- 86954 2000/08/18 05:46:26 2.2_rev0_CDa/sparc/binary-sparc-1.list -rw-r--r-- 679356416 2000/08/18 05:27:40 2.2_rev0_CDa/sparc/binary-sparc-1_NONUS.iso -rw-r--r-- 90887 2000/08/18 05:47:42 2.2_rev0_CDa/sparc/binary-sparc-1_NONUS.list -rw-r--r-- 668405760 2000/08/18 05:29:12 2.2_rev0_CDa/sparc/binary-sparc-2.iso -rw-r--r-- 123413 2000/08/18 05:50:24 2.2_rev0_CDa/sparc/binary-sparc-2.list -rw-r--r-- 314284032 2000/08/18 05:29:57 2.2_rev0_CDa/sparc/binary-sparc-3.iso -rw-r--r-- 41686 2000/08/18 05:51:55 2.2_rev0_CDa/sparc/binary-sparc-3.list % -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Two more corrupt .debs found from 2.2_rev0 images
After checking all the CD images for all the architectures, I found two more corrupt files. The sparc images I tested were 2.2_rev0_CDa versions, the old 2.2_rev0 images were not tested. Testing included _only_ .deb files Here's what I found. binary-powerpc-1_NONUS.iso == dists/potato/main/binary-powerpc/otherosfs/smbfs_2.0.7-3.deb 2e2c1d222b3cdb9db70ebf09634994c2, correct is 92725d3d06262927fee42029c308f84e binary-powerpc-3.iso dists/potato/main/binary-all/doc/gimp-manual_1.0.0-4.deb 5824b918062f3ad9d9a2b08c97b78634, correct is f01a6b0eb32c8dc38cc5f52067eb2ae9 Here is the perl script I used. #!/usr/bin/perl if ($ARGV[0] eq "-v") { $verbose = 1; shift @ARGV; } if ($#ARGV < 0) { print STDERR "usage: $0 [-v] md5sums [md5sums ...]\n"; print STDERR "Calculates MD5 checksums for .deb files in current and every subdirectory.\n"; print STDERR " These checksums are compared against checksums in md5sums files.\n"; print STDERR " md5sums files are typically md5sums files from debian/indices/ "; print STDERR " and debian-non-US/indices-non-US/\n"; print STDERR "\n"; print STDERR " -vList also files that were not found from md5sums files,\n"; print STDERR " and MD5 checksum failures with non .deb files.\n"; exit 1; } # Typical usage goes like this: # 1. Make directories called images, sums and loop # 2. Put the CD ISO images in the images directory # 3. Put the MD5 checksum files debian/indices/md5sums and # debian-non-US/indices-non-US/md5sums into sums directory # 4. for each $image in images/* do # 5. sudo mount -o loop $image loop # 6. cd loop # 7. check-md5.pl ../sums/* # 8.Hope that there's no output # 9. cd .. # 10. sudo umount loop while ($sumfile = shift @ARGV) { open MD5SUMS, "< $sumfile" or die "$sumfile: $!\n"; while () { s|dists/sid|dists/potato|; # CD images do not have sid. They have potato ($md5, $file) = (/^(\S+)\s+(\S+)$/); $sums{$file} = $md5; } } open FILES, "find . -type f |" or die "find failed: $!"; while () { s|./||; next if (/TRANS.TBL$/); $line = `md5sum $_`; ($md5, $file) = ($line =~ /^(\S+)\s+(\S+)$/); $correct_md5 = $sums{$file}; if ($verbose and not defined $correct_md5) { print "$file not listed in md5sums files\n"; } elsif ($md5 != $correct_md5) { next if not $verbose and not ($file =~ /\.deb$/); print "$file $md5, correct is $correct_md5\n"; } } // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Corrupt file on 2.2_rev0/i386/binary-i386-2.iso ?
Philip Hands <[EMAIL PROTECTED]> wrote: > Could someone please mount the CDs on a machine with a known good FTP > archive, and compare the CD files with those in the archive, to > establish the extent of the problem? I thought I would do the following: 1. Mount an image over loopback 2. Calculate MD5 checksums for each file in the image 3. Get the md5sums file from debian/indices/ 4. Compare the MD5 checksum listings generated in step 2 and 3 Does this sound reasonable? I can start with alpha and arm disks immediately. // Heikki -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Corrupt file on 2.2_rev0/i386/binary-i386-2.iso ?
John Pearson <[EMAIL PROTECTED]> wrote: > >From my local mirror: > # md5sum /cognac/linux/debian/dists/potato/main/binary-i386/shells/pdksh_5.2.14-1.deb > 74ee8a1d7c6655c8592994c5515929a1 >/cognac/linux/debian/dists/potato/main/binary-i386/shells/pdksh_5.2.14-1.deb > Works for me. Last rsync'd against ftp.debian.org after 2.2 was announced. Hmm, maybe I did not express myself clearly enough, but I am not claiming that pdksh_5.2.14-1.deb in the mirrors is corrupt. The problem is with pdksh_5.2.14-1.deb that is in the CD ISO image binary-i386-2.iso. If you have made CDs from the CD images and install pdksh from there, does it still work? // Heikki -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Corrupt file on 2.2_rev0/i386/binary-i386-2.iso ?
I received a report from a person that dists/potato/main/binary-i386/shells/pdksh_5.2.14-1.deb might be corrupt on binary-i386-2.iso. The ISO image matches its MD5SUM so the image should be correct. The transcript below shows that I can mount the image using the loopfs, but data.tar.gz inside pdksh_5.2.14-1.deb is corrupt. The MD5 checksum 74ee8a1d7c6655c8592994c5515929a1, which is on the ISO image's md5sum.txt file, matches what I get from the real filesystem. Can someone verify if the file is actually corrupt, or is it just the local image on trumpetti.atm.tut.fi that is incorrect. % pwd /export/opt/cdtemp/loop2 % mount |grep loop2 /export/opt/cdtemp/binary-i386-2.iso on /export/opt/cdtemp/loop2 type iso9660 (rw,loop=/dev/loop1) % % md5sum dists/potato/main/binary-i386/shells/pdksh_5.2.14-1.deb 882b2cde3a362198f73c599aa7fabbb8 dists/potato/main/binary-i386/shells/pdksh_5.2.14-1.deb % % grep dists/potato/main/binary-i386/shells/pdksh_5.2.14-1.deb md5sum.txt 74ee8a1d7c6655c8592994c5515929a1 ./dists/potato/main/binary-i386/shells/pdksh_5.2.14-1.deb % % mkdir /tmp/foo % cp dists/potato/main/binary-i386/shells/pdksh_5.2.14-1.deb /tmp/foo % cd /tmp/foo % ar x pdksh_5.2.14-1.deb % ls -l total 429 -rw-r--r--1 hessuhessu 641 Aug 18 11:52 control.tar.gz -rw-r--r--1 hessuhessu 216576 Aug 18 11:52 data.tar.gz -rw-r--r--1 hessuhessu 4 Aug 18 11:52 debian-binary -r--r--r--1 hessuhessu 217410 Aug 18 11:52 pdksh_5.2.14-1.deb % tar -ztvf data.tar.gz drwxr-xr-x root/root 0 1999-09-25 00:26 ./ drwxr-xr-x root/root 0 1999-09-25 00:26 usr/ drwxr-xr-x root/root 0 1999-09-25 00:26 usr/doc/ drwxr-xr-x root/root 0 1999-09-25 00:26 usr/doc/pdksh/ -rw-r--r-- root/root 3607 1999-09-25 00:26 usr/doc/pdksh/copyright -rw-r--r-- root/root 48193 1999-09-25 00:26 usr/doc/pdksh/changelog.0.gz -rw-r--r-- root/root 18376 1999-09-25 00:26 usr/doc/pdksh/changelog.gz -rw-r--r-- root/root 10681 1999-09-25 00:26 usr/doc/pdksh/NEWS.gz -rw-r--r-- root/root 7833 1999-09-25 00:26 usr/doc/pdksh/NOTES.gz -rw-r--r-- root/root 920 1999-09-25 00:26 usr/doc/pdksh/changelog.Debian.gz drwxr-xr-x root/root 0 1999-09-25 00:26 usr/bin/ -rwxr-xr-x root/root162448 1999-09-25 00:26 usr/bin/ksh drwxr-xr-x root/root 0 1999-09-25 00:26 usr/man/ drwxr-xr-x root/root 0 1999-09-25 00:26 usr/man/man1/ -rw-r--r-- root/root 41805 1999-09-25 00:26 usr/man/man1/ksh.1.gz gzip: stdin: invalid compressed data--crc error tar: Child returned status 1 tar: Error exit delayed from previous errors zsh: exit 2 tar -ztvf data.tar.gz % -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: New fast, soon complete mirror.
Mattias Wadenstein <[EMAIL PROTECTED]> wrote: > We now have all the images. The md5sums have been verified. Thanks for the pointer, trumpetti.atm.tut.fi also has a complete set of images now. > Especially those within nordunet should seriously consider syncing from us > now. Getting the last NONUS arm image was very fast. It only took a couple of minutes :-) > There seems to be some problem with the connection between > cdimage.debian.org and nordunet, some of the NONUS isos we downloaded from > ftp.gigabell.net and some took an extra roundtrip via another system with > other connectivity to the world. Sounds familiar. That last time c.d.o worked well for us was on Monday. Since that the connection has been very bad. I wonder how widespread the problem is. // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Best mirror for cd imnage rsync?
Nathan E Norman <[EMAIL PROTECTED]> wrote: > Anyone care to toss me a clue? I want images but I don't want to > hose c.d.o or use unnecessary bandwidth! Try trumpetti.atm.tut.fi. It is identical to c.d.o except that binary-arm-1_NONUS.iso is still missing. There is something between us and c.d.o which makes transfers extremely slow and prone to hang ups. // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Full set of 2.2_rev0 sparc images now available
The sparc images are now available at trumpetti.atm.tut.fi. Thanks goes to Jonathan Quick from .za. // Heikki -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: rsync - binary-i386-3.iso
Mattias Wadenstein <[EMAIL PROTECTED]> > I'm currently working on the other architectures, trumpetti.atm.tut.fi has > most but not the NONUS images among the other architectures. And neither > has any mirror I have found in the rsync-mirror-listing. trumpetti.atm.tut.fi currently has arm, i386, m68k, powerpc and source available. Only i386 and source have NONUS images, which I managed to get on Monday, when they had just arrived at c.d.o. I am currently trying to rsync my arm, m68k and powerpc TC3 NONUS images to 2.2_rev0 images but have had no luck with c.d.o. It's either full or I get < 1MB and then the connection times out. I even downgraded my rsync from 2.4.2 to 2.3.1 but it made no difference. If anyone has the NONUS images available, please speak up :-) Also, has anyone managed to get the alpha or sparce images? // Heikki -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Any success mirroring 2.2_rev0?
jason andrade <[EMAIL PROTECTED]> wrote: > i can imagine it being slow - rsync is definitely more of a cpu suck > than ftp (mirror) as well as the usual mirror bandwidth issues. I can feel its pain. A little while ago I posted a pointer to the main Finnish Linux newsgroup and the interface utilisation measured from the switch is now about 35-45%. The interface is 100Mbit/s full-duplex ethernet :-) // Heikki -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Any success mirroring 2.2_rev0?
Has anyone managed to mirror the other archs besides i386 and source from cdimage.debian.org? I got those two yesterday afternoon, but since then cdimage.debian.org has been _really_ slow and rsyncs have just stalled after initially transferring a small amount of data. If I just let it sit there, it usually exists after timeout. If you are still trying to get i386 or source images, give trumpetti.atm.tut.fi a try. We have a 2 cpu box with fast net connection dedicated to mirroring things. // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: 2.2 CD images appearing
Philip Hands <[EMAIL PROTECTED]> wrote: > The 2.2 CD images are available from cdimage.debian.org. > If people could mention when they've got copies of the images, so that > others can start mirroring from them, it might help spread the load > more (cdimage.d.o is only a P166, so cannot stand too many rsyncs). trumpetti.atm.tut.fi currently has i386 and source images. FTP: ftp://trumpetti.atm.tut.fi/debian-cd/ HTTP: http://trumpetti.atm.tut.fi/debian-cd/ Rsync: trumpetti.atm.tut.fi::debian-cd/ The images were rsynced on top of the existing TC3 images with minimal bandwith usage. More will be added once I manage to get them. If you decide to use rsync, please remeber to use -H ( --hard-links ) or you get everything in double. Phil, only source/MD5SUMS had PGP signature, is this ok? // Heikki -- Heikki Vatiainen * [EMAIL PROTECTED] Tampere University of Technology * Tampere, Finland -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]