Bug 75377 (pyschediscinfo)
| Summary: | Hitting a bad package can lead to a traceback | ||||||
|---|---|---|---|---|---|---|---|
| Product: | [Retired] Red Hat Linux | Reporter: | Jeremy Katz <katzj> | ||||
| Component: | anaconda | Assignee: | Jeremy Katz <katzj> | ||||
| Status: | CLOSED RAWHIDE | QA Contact: | Brock Organ <borgan> | ||||
| Severity: | medium | Docs Contact: | |||||
| Priority: | medium | ||||||
| Version: | 8.0 | CC: | 2pascal, administrator, aguzman, a__j__f, beacon, bellrs, bennekommm, ben, bryangaston122, c5engnr, camerons, cefalany, charles_g82, chyongfeng, david, desmoh, fadertg, fadi_badine, fatihb, gareth_alexander, gkuzet, hkr, info, info, itsmerico, j.h.bouwsma, jmerli, jmoran, junglist_bunnies, kell_ja, kmichaud, larryd2756, lg5000, linear8002, liy, manda73, manu, mgmg, michael, mil, mnelson, ozgenbulut, pat, powerochezz, prhodes, raegan, robertwilson, rumors, salil6, sbhennessy, sherbin2, shishz, smcleod, srodgers, stormbringer49, trrammler, trysnat, tschiffer, vivekkhanna, vkarasik | ||||
| Target Milestone: | --- | ||||||
| Target Release: | --- | ||||||
| Hardware: | i386 | ||||||
| OS: | Linux | ||||||
| Whiteboard: | |||||||
| Fixed In Version: | Doc Type: | Bug Fix | |||||
| Doc Text: | Story Points: | --- | |||||
| Clone Of: | Environment: | ||||||
| Last Closed: | 2002-10-24 18:55:34 UTC | Type: | --- | ||||
| Regression: | --- | Mount Type: | --- | ||||
| Documentation: | --- | CRM: | |||||
| Verified Versions: | Category: | --- | |||||
| oVirt Team: | --- | RHEL 7.3 requirements from Atomic Host: | |||||
| Cloudforms Team: | --- | Target Upstream Version: | |||||
| Embargoed: | |||||||
| Attachments: |
|
||||||
|
Description
Jeremy Katz
2002-10-07 20:42:59 UTC
*** Bug 75246 has been marked as a duplicate of this bug. *** *** Bug 74718 has been marked as a duplicate of this bug. *** *** Bug 75258 has been marked as a duplicate of this bug. *** *** Bug 75263 has been marked as a duplicate of this bug. *** *** Bug 75310 has been marked as a duplicate of this bug. *** *** Bug 75360 has been marked as a duplicate of this bug. *** I have also been able to replicate this on the 7.3 iso's. I had to go back to 7.1 to get an install. Next, I tried an upgrade to 8.0 which also failed. So, now I'm trying the upgrade to 7.3. Which has subsequently failed....... Actually, the failure case under 7.3 is different. And in fixing it, it uncovered something else. Unfortunately, my CD-ROM drives were correcting errors before they got to the level where this code would get triggered when I tested the fix :/ An update disk for Red Hat Linux 8.0 is available at http://people.redhat.com/~katzj/8.0-cdread.img that should correct this problem. If you download this image and place it on a floppy disk (using dd or rawrite as with a boot disk), and then boot with 'linux updates' from the syslinux prompt, you will be prompted to provide this disk. *** Bug 75201 has been marked as a duplicate of this bug. *** *** Bug 75227 has been marked as a duplicate of this bug. *** I tried the 8.0-cdread.img file. Same read error. However, this time when I hit "okay" the cd was ejected and I was prompted to insert disk 1. I pushed the cd back in, it remounted and the install continued. This happened 7 times on Disk 1. Disk 2 and 3 did not have the any problems. As it stands now I have a fully working machine. Thanks for the quick fix. So this image corrects bug I mentioned as no.1 from bug #74718 - anaconda not able handle situation. Now left to find why CD is unmounted, mount directory deleted and device is changing from hdc to cdrom and vice versa. Tried your fix. Rawrite won't find a file not in 8.3 format "8.0-cdread.img" has 10 characters in front of the extension. Understand that I'm a complete newbie at Linux, so forgive me if I don't know something obvious. Even tried renaming the file to "80cdread.img" and did a rawrite to a floppy, but once I ran the linux updates command anaconda started up and received a "signal 11" and aborted. What am I doing wrong here? *** Bug 75201 has been marked as a duplicate of this bug. *** *** Bug 75441 has been marked as a duplicate of this bug. *** *** Bug 75475 has been marked as a duplicate of this bug. *** Rather than rawwrite use rawwritewin. Both are located in the dosutils directory under RH Linux 8.0 on the ftp sites. Besides rawwritewin, download also the one dll file. The rawwritewin app is more compatible with the later windows releases, imho. Its straightforward. Haven't had the chance to apply the path but I do have the diskette readied. *** Bug 75559 has been marked as a duplicate of this bug. *** *** Bug 75848 has been marked as a duplicate of this bug. *** *** Bug 75837 has been marked as a duplicate of this bug. *** *** Bug 75775 has been marked as a duplicate of this bug. *** *** Bug 75766 has been marked as a duplicate of this bug. *** Burning the ISO to disk and then installing from CD seems to work best. I didn't have any problems with the installation that way. *** Bug 75721 has been marked as a duplicate of this bug. *** When I tried to save the http://people.redhat.com/~katzj/8.0-cdread.img file in a floppy (1.44MB) I got into disk space problem as the file is > 1.44MB. I think I am not clear about the steps: I am not able to install RH8.0 yet (has Win98 in another partition) 1. Can I download the file in my hard drive and then use some tool (dd?, don't know how to use it) to move the file in a floppy. 2. Is dd or rawrite linux tools. How to use it? *** Bug 75962 has been marked as a duplicate of this bug. *** *** Bug 75921 has been marked as a duplicate of this bug. *** *** Bug 75929 has been marked as a duplicate of this bug. *** *** Bug 76038 has been marked as a duplicate of this bug. *** *** Bug 76045 has been marked as a duplicate of this bug. *** I had most of the same problems listed. The cdread.img helped except that there are some packages that result in a "fatal" error and as such do not allow you the opportunity of a retry. If they did, I think I might have made it. This is a serious problem. I've tried a couple of different images from different mirrors and I don't think it is a problem with the download. The installs are being attempted on DELL GX150's and this is a new machine. It did the same on the older Dells we have too. This still doesn't work for me. I "only" tried 3 times, so I'll go on trying, but it gives, after re-insertion of the first CD, the same error. *** Bug 76249 has been marked as a duplicate of this bug. *** *** Bug 76299 has been marked as a duplicate of this bug. *** *** Bug 75093 has been marked as a duplicate of this bug. *** *** Bug 76404 has been marked as a duplicate of this bug. *** Tried this weekend with generic AMD 500Mhz machine with the SAME CD's that I tried with my Dell's (INTEL). Install worked flawlessly without so much as a hiccup. This "feels" like something is going on with the CD drives/error correction or something related. Either that or the Dell bioses have something unusual in them that is corrupting things. On both Dell's I get "bad magic - cpio" messages at the very start just before the linux install prompt appears indicating something wrong with the CD images apparently. Yet these same CD's load other machines just fine. Until I got a flawless install on the generic AMD box, I suspected my downloads (even though the media checks are good.) Now I suspect something in the Dell architecture/devices. I had the same problem on a shuttle AI61 board using a generic ide-cdrom its a slot-A board that uses the AMD chipset. Its not just a dell bios problem. My mitsumi cd-rw drive on my other box says the media is ok when you do the media check. But when i test the same disc on my athlon system, ive gotten both good and bad results, but it always locksup during the install or does the fatal package error thing. My other suspicion is the video hardware or something to do with that. The tranceback messages/packages seem to have a video "flavor/theme" to them if the naming conventions mean anything. (gui.py and the like.) Just a hunch. I may try a different CDROM, taking the one out of the "flawless" machine and putting it into the Dell, just to eliminate one variable. The Dell's have on board video so I can't change the video easily. *** Bug 76585 has been marked as a duplicate of this bug. *** *** Bug 76413 has been marked as a duplicate of this bug. *** *** Bug 76412 has been marked as a duplicate of this bug. *** *** Bug 76611 has been marked as a duplicate of this bug. *** *** Bug 76405 has been marked as a duplicate of this bug. *** I tried the update disk and I still have this problem installing 8.0. Every time it gets to zlip, it fails. Hitting OK at the traceback results in a reboot. Created attachment 82664 [details]
Blocco lettura RPM
I tried making the floppy in order to get some kind of debugging info. It did make a floppy, but it seems wrong. It contains only the following files: [rutt@mojo floppy]$ ls boot.msg general.msg ldlinux.sys rescue.msg syslinux.cfg vmlinuz expert.msg initrd.img param.msg snake.msg template.img Is one of these the right thing? Anyway, I'm dead in the water at the moment. Using the updates floppy as suggested in the fix message seems to do nothing. I'd suspect that I have a bad cd, but the problem is repeatable and also occurs exactly with the zlib install as reported in earlier reports, so this doesn't really seem likely. Please help soon. After a closer reading of the comments above I tried swapping cd drives and had no problem during the installation. The problem cd drive was an Acer cdrw 1208a--002. Perhaps this info will help someone else. *** Bug 76818 has been marked as a duplicate of this bug. *** *** Bug 76994 has been marked as a duplicate of this bug. *** *** Bug 76988 has been marked as a duplicate of this bug. *** *** Bug 76864 has been marked as a duplicate of this bug. *** *** Bug 76864 has been marked as a duplicate of this bug. *** *** Bug 76784 has been marked as a duplicate of this bug. *** *** Bug 76850 has been marked as a duplicate of this bug. *** *** Bug 76850 has been marked as a duplicate of this bug. *** Still I see problem. This bug says: "If you hit a case where a package can't be read and you get prompted to try again, you can end up in a case where the CD isn't mounted, so trying to read the .discinfo file fails." OK, but why installer is missing package? In my opinion, this happens in reversed order than bug is saiing - means first /tmp/hdc disappears, then package is corrupted, and then also .discinfo is not found. So .img solves only traceback problem of anaconda not the roots... *** Bug 77033 has been marked as a duplicate of this bug. *** *** Bug 77261 has been marked as a duplicate of this bug. *** *** Bug 77257 has been marked as a duplicate of this bug. *** *** Bug 77141 has been marked as a duplicate of this bug. *** *** Bug 77321 has been marked as a duplicate of this bug. *** *** Bug 75456 has been marked as a duplicate of this bug. *** *** Bug 77595 has been marked as a duplicate of this bug. *** FWIW - I've done at least two other installs on an IBM 300P (Pentium III) and an old AMD 233-K6 machine and the SAME CD's that blew up on my Dells had no trouble whatsoever with the install. *** Bug 77891 has been marked as a duplicate of this bug. *** *** Bug 77891 has been marked as a duplicate of this bug. *** *** Bug 77891 has been marked as a duplicate of this bug. *** *** Bug 77819 has been marked as a duplicate of this bug. *** *** Bug 77925 has been marked as a duplicate of this bug. *** *** Bug 78030 has been marked as a duplicate of this bug. *** *** Bug 78012 has been marked as a duplicate of this bug. *** Guys, Forgive my ignorance, but like some of the other comments made here, I have no idea how to execute the procedures written to resolve this error and install RedHat 8.0 on my PC. I think there is too much Linux knowledge assumed here. I have a goal which is to find something better than a Windows platform and I was really hoping RedHat Linux was the way to go. So far I have 6 impressive CD's which I bought and no Linux on my PC. Can someone "in the know" list a set of steps of how to resolve this problem. I don't even care if it is just a work-a-round, as long as I can get RedHat 8.0 installed on my PC. Fitting files greater than 1.44 on to a floppy would be a good starting point. Thanks in advance for the copius amounts of help coming my way. *** Bug 78154 has been marked as a duplicate of this bug. *** *** Bug 78136 has been marked as a duplicate of this bug. *** *** Bug 77931 has been marked as a duplicate of this bug. *** *** Bug 77879 has been marked as a duplicate of this bug. *** *** Bug 77566 has been marked as a duplicate of this bug. *** *** Bug 78315 has been marked as a duplicate of this bug. *** *** Bug 78208 has been marked as a duplicate of this bug. *** *** Bug 78525 has been marked as a duplicate of this bug. *** *** Bug 78476 has been marked as a duplicate of this bug. *** *** Bug 78585 has been marked as a duplicate of this bug. *** *** Bug 78454 has been marked as a duplicate of this bug. *** I added this comment to Bug 78656 (which appears to be a duplicate of this bug). The reporter seemed a bit confused on the steps necessary to apply the image to a floppy. Hopefully my comments will help out (please feel free to amend the steps if I left out something). The boot portion of the Red Hat Linux 8.0 installation disc and Anaconda, the Red Hat Linux installer, were designed with the philosophy that even the "gold" edition could have bugs that hinder Red Hat Linux being installed on certain systems. The .discinfo problem seems to be sort of a timing issue - specifically related to certain CD-ROM drives (I might be over-simplifying the problem). Nonetheless, people filed the bug in Bugzilla - the Red Hat engineers went to work to create an update image. When you first put in the Red Hat Linux 8.0 disc 1, it shows you a text screen showing some options. Typing "linux updates" at the prompt let's the installation know that you have a disk that contains some patches or workarounds to get rid of a problem. You will then be asked to provide the disk that you wrote the image on. See the following steps: 1. Most people have access to a Windows machine - if you do, you should first obtain the rawwritewin application. You can find the rawwritewin application at the following address: ftp://mirror.hiwaay.net/redhat/redhat/linux/8.0/en/os/i386/dosutils/rawritewin/ If you have access to a Linux box, you can also use "dd" - go to step 2. Download all the files in the directory to your local computer. "rawwritewin" will take an image (like the image file mentioned in the bug report) and byte-for-byte write it to your floppy disk. 2. You should then download the actual image to be written to the floppy disk. http://people.redhat.com/~katzj/8.0-cdread.img - place the 8.0-cdread.img file in a convenient location. 3. Put a 1.44M floppy disk in your drive - make sure you don't mind losing everything on the disk! 4. Windows - Double-click "rawwritewin.exe" (where ever you put the file) - it will pop-up a dialog box that says "Write", "Read", etc. in the tabs. The Floppy Drive selection box should already be set to your default A: drive. Click the "..." button to the right of the Image file: text box. Now, find the image that you downloaded - Click OK. Now click the "Write" button in the lower right of your tab area. This will byte-for-byte write the update disk to your floppy. Linux - if you have the image in a convenient spot - you can write the image as such: dd if=<path to image file>/8.0-cdread.img of=/dev/fd0 bs=1440k 5. Now you are ready to install Red Hat Linux 8.0 again! 6. Boot off disc 1. 7. Text will appear (the SYSLINUX menu) - at the prompt type: linux updates 8. Follow the instructions and you should be good-to-go. I hope this helps you out. *** Bug 78712 has been marked as a duplicate of this bug. *** *** Bug 78853 has been marked as a duplicate of this bug. *** *** Bug 78797 has been marked as a duplicate of this bug. *** *** Bug 78757 has been marked as a duplicate of this bug. *** *** Bug 78656 has been marked as a duplicate of this bug. *** *** Bug 78869 has been marked as a duplicate of this bug. *** *** Bug 79019 has been marked as a duplicate of this bug. *** *** Bug 78985 has been marked as a duplicate of this bug. *** *** Bug 79265 has been marked as a duplicate of this bug. *** hmm...i get a problem like this. it cant install slang-1.4.5-11.i286.rpm i think this is a problem that happends every day at 1000's of people. and its a big - for Red Hat. anyway, ill try what michael wrote above me, dont know if i can do it. alltough its maybe a perfect answer. im not english speaking, so its more difficult to me. I have been informed by quite a few people that my plain-explanation of how to apply the above fix has been helpful. There are some, however, that still have problems installing Red Hat Linux 8.0 even after applying katzj's update. After investigating one-on-one with various people having the ".discinfo" problem - I've come up with the following conclusion which I added to Bug 79369. I am beginning to think that there are two actual problems - though one is dependent on the vast number of others. First problem - Anaconda that is shipped with Red Hat Linux 8.0 does not properly handle the inability to read packages that appear to be "corrupt" (to some degree). This is largely corrected by the procedures listed in bug 75377. Second problem - ailing CD-ROM drives (be it a dirty lens, firmware issues, or a failing drive altogether) seem to trigger this bug. You have properly pointed out that it installs without a hitch on your laptop (thus, the CDs themselves are good - which means the stock shipment of Anaconda never encounters the second problem [which triggers the first problem]). I would suggest the following: 1) Replace the EIDE cable from the motherboard that attaches to your CD-ROM drive with a fresh one. Bad cables are notorious for spurring strange errors (seek and read). Install again. If it installs properly - the culprit seems to be the faulty cable. If not, proceed to step two. 2) Move your CD-ROM drive to a different EIDE interface (for example, from the secondary channel master to the primary channel slave). Install again. If it installs properly, then there is a quirk between the interface of your CD-ROM and the EIDE controller. This could be due to a variety of factors - namely: (1) failing CD-ROM chipset - I've had a hard drive that would work only as a slave. (2) failing EIDE controller on the motherboard - I've had a channel go flaky on a motherboard and then quit functioning altogether. If it does not install properly, proceed to step three. 3) Replace your CD-ROM drive with a new one (or one that has successfully installed Red Hat Linux 8.0 on a machine). If it installs properly, your original CD-ROM drive may need to be cleaned or it is out of calibration (which should never happen, but if it does it means CD-ROM drive death). By step 3 you should be able to install Red Hat Linux 8.0 without a hitch. Even the Windows installer does not handle ailing CD-ROM drives (or EIDE interfaces) gracefully. Hardware problems are assumed to NOT exist in software (generally). If it doesn't install properly - you've hit something particularly strange - which could me costly repairs (like motherboard replacement). Fluctuations in voltage on a motherboard can also cause VERY strange problems that trickle down to the application level. Conclusion: Although there appears to be a wide-spread problem with the Anaconda that shipped with Red Hat Linux 8.0, I do not believe that to be the case (necessarily). Hardware issues (quirks and errors) cause erratic behaviour in software. Microsoft also experienced a problem of a greater magnitude when thousands of consumers complained that installing Windows 95 was corrupting there machines after a reboot. In later service releases of Windows 95, Microsoft required Scandisk to check for strange anomalies on the file system, and more specifically the hard drive itself. The solution was not to engineer the installer to ignore such strange errors - but to inform the user that the hardware in the system was failing and needed to be repaired or replaced before continuing with the install process. The Anaconda installer that ships with Red Hat Linux 8.0 is pretty much sound as far as being "stable" - however, the quirks that can arise from "well-worn" hardware can trigger a problem based on an assumption in Anaconda (which has been properly patched). My recommendation to the Anaconda team at Red Hat is to also add a recommendation to the user that hardware may be failing if this behaviour continues even with the fixes applied to Anaconda. The test of this theory is to actually botch a CD-ROM drive or two (applying a finger print or dust on the lens of a CD-ROM drive) and see how Anaconda handles such hardware. Also, ailing computers should not be retired from the labs at Red Hat - I would imagine that most hodge-podge computers out in the world that would potentially be running Red Hat Linux have some sort of quirk (stemming from ailing hardware). Being able to handle as such would be useful. *** Bug 79614 has been marked as a duplicate of this bug. *** *** Bug 79628 has been marked as a duplicate of this bug. *** *** Bug 79369 has been marked as a duplicate of this bug. *** *** Bug 79908 has been marked as a duplicate of this bug. *** Time tracking values updated *** Bug 80107 has been marked as a duplicate of this bug. *** *** Bug 80630 has been marked as a duplicate of this bug. *** *** Bug 80637 has been marked as a duplicate of this bug. *** *** Bug 81337 has been marked as a duplicate of this bug. *** *** Bug 81333 has been marked as a duplicate of this bug. *** *** Bug 81088 has been marked as a duplicate of this bug. *** *** Bug 81376 has been marked as a duplicate of this bug. *** *** Bug 81274 has been marked as a duplicate of this bug. *** *** Bug 78847 has been marked as a duplicate of this bug. *** *** Bug 78226 has been marked as a duplicate of this bug. *** *** Bug 81442 has been marked as a duplicate of this bug. *** *** Bug 79741 has been marked as a duplicate of this bug. *** *** Bug 78833 has been marked as a duplicate of this bug. *** *** Bug 78714 has been marked as a duplicate of this bug. *** *** Bug 71269 has been marked as a duplicate of this bug. *** The fix helped me out. Thanx I guess I have to comment in order to stop the emails. As a newbie, I didn't know how to check the other bugs. With that in mind, I followed the directions on bug 75377, and Vi-ola, things work just famously! Thank you all for your prompt support. Pat *** Bug 81616 has been marked as a duplicate of this bug. *** *** Bug 82154 has been marked as a duplicate of this bug. *** thx for the fix *** Bug 82335 has been marked as a duplicate of this bug. *** *** Bug 82775 has been marked as a duplicate of this bug. *** *** Bug 82908 has been marked as a duplicate of this bug. *** *** Bug 83083 has been marked as a duplicate of this bug. *** *** Bug 83093 has been marked as a duplicate of this bug. *** *** Bug 83256 has been marked as a duplicate of this bug. *** *** Bug 83275 has been marked as a duplicate of this bug. *** *** Bug 83346 has been marked as a duplicate of this bug. *** *** Bug 83465 has been marked as a duplicate of this bug. *** *** Bug 83796 has been marked as a duplicate of this bug. *** *** Bug 83804 has been marked as a duplicate of this bug. *** *** Bug 83831 has been marked as a duplicate of this bug. *** *** Bug 84050 has been marked as a duplicate of this bug. *** *** Bug 83831 has been marked as a duplicate of this bug. *** *** Bug 83831 has been marked as a duplicate of this bug. *** *** Bug 84563 has been marked as a duplicate of this bug. *** *** Bug 79432 has been marked as a duplicate of this bug. *** *** Bug 84895 has been marked as a duplicate of this bug. *** *** Bug 84899 has been marked as a duplicate of this bug. *** *** Bug 85438 has been marked as a duplicate of this bug. *** *** Bug 85383 has been marked as a duplicate of this bug. *** *** Bug 85412 has been marked as a duplicate of this bug. *** *** Bug 85693 has been marked as a duplicate of this bug. *** *** Bug 75197 has been marked as a duplicate of this bug. *** *** Bug 86169 has been marked as a duplicate of this bug. *** *** Bug 86348 has been marked as a duplicate of this bug. *** *** Bug 86462 has been marked as a duplicate of this bug. *** *** Bug 87387 has been marked as a duplicate of this bug. *** *** Bug 88694 has been marked as a duplicate of this bug. *** *** Bug 89095 has been marked as a duplicate of this bug. *** *** Bug 89550 has been marked as a duplicate of this bug. *** *** Bug 90813 has been marked as a duplicate of this bug. *** *** Bug 91609 has been marked as a duplicate of this bug. *** *** Bug 91614 has been marked as a duplicate of this bug. *** *** Bug 97025 has been marked as a duplicate of this bug. *** *** Bug 97429 has been marked as a duplicate of this bug. *** *** Bug 97882 has been marked as a duplicate of this bug. *** *** Bug 97993 has been marked as a duplicate of this bug. *** *** Bug 98711 has been marked as a duplicate of this bug. *** *** Bug 101144 has been marked as a duplicate of this bug. *** *** Bug 103143 has been marked as a duplicate of this bug. *** *** Bug 104487 has been marked as a duplicate of this bug. *** *** Bug 104590 has been marked as a duplicate of this bug. *** *** Bug 110559 has been marked as a duplicate of this bug. *** *** Bug 115532 has been marked as a duplicate of this bug. *** *** Bug 117811 has been marked as a duplicate of this bug. *** |