Hi,
I'm trying to install Centos 5.2 on an "old" machine in graphical mode (to use LVM) but anaconda crash. I made several tries with various installation resquest (the last is "Server", "Customize after installation") unfortunately, it's reproductible:
Traceback (most recent call first): File "/usr/lib/python2.4/site-packages/yum/sqlitesack.py", line 42, in newFunc raise Errors.RepoError File "/usr/lib/python2.4/site-packages/yum/sqlitesack.py", line 596, in searchPrco results.extend(self.searchFiles(name)) File "/usr/lib/python2.4/site-packages/yum/sqlitesack.py", line 40, in newFunc return func(*args, **kwargs) File "/usr/lib/python2.4/site-packages/yum/sqlitesack.py", line 648, in searchProvides return self.searchPrco(name, "provides") File "/usr/lib/python2.4/site-packages/yum/packageSack.py", line 399, in _computeAggregateListResult sackResult = apply(method, args) File "/usr/lib/python2.4/site-packages/yum/packageSack.py", line 267, in searchProvides return self._computeAggregateListResult("searchProvides", name) File "/usr/lib/python2.4/site-packages/yum/depsolve.py", line 149, in whatProvides pkgs = self.pkgSack.searchProvides(name) File "/tmp/treedir.8196/instimage/usr/lib/anaconda/yuminstall.py", line 388, in _provideToPkg for po in self.whatProvides(r, f, v): File "/tmp/treedir.8196/instimage/usr/lib/anaconda/yuminstall.py", line 461, in tsCheck dep = self._provideToPkg(req) File "/tmp/treedir.8196/instimage/usr/lib/anaconda/yuminstall.py", line 429, in resolveDeps unresolved = self.tsCheck(unresolved) File "/usr/lib/python2.4/site-packages/yum/__init__.py", line 585, in buildTransaction (rescode, restring) = self.resolveDeps() File "/tmp/treedir.8196/instimage/usr/lib/anaconda/yuminstall.py", line 1265, in doPostSelection (code, msgs) = self.ayum.buildTransaction() File "/tmp/treedir.8196/instimage/usr/lib/anaconda/backend.py", line 177, in doPostSelection return anaconda.backend.doPostSelection(anaconda) File "/usr/lib/anaconda/dispatch.py", line 201, in moveStep rc = stepFunc(self.anaconda) File "/usr/lib/anaconda/dispatch.py", line 124, in gotoNext self.moveStep() File "/usr/lib/anaconda/gui.py", line 1018, in nextClicked self.anaconda.dispatch.gotoNext() RepoError: None
Can somebody tell me where is the problem? Should I have to report a bug? On Centos Mantis? On RH Bugzilla?
Regards, Alain
On Tue, Dec 30, 2008 at 10:28 AM, Alain PORTAL alain.portal@free.fr wrote:
Hi,
I'm trying to install Centos 5.2 on an "old" machine in graphical mode (to use LVM) but anaconda crash. I made several tries with various installation resquest (the last is "Server", "Customize after installation") unfortunately, it's reproductible:
You need to supply more information about this "old" machine - that could be the problem....
mhr
Le mardi 30 décembre 2008, MHR a écrit :
You need to supply more information about this "old" machine - that could be the problem....
What do you need as information? I'm not in from of this machine tonight, the only thing I remember is it has 512MB RAM and 40GB HDD.
Regards, Alain
On Tue, Dec 30, 2008 at 2:03 PM, Alain PORTAL alain.portal@free.fr wrote:
Le mardi 30 décembre 2008, MHR a écrit :
You need to supply more information about this "old" machine - that could be the problem....
What do you need as information? I'm not in from of this machine tonight, the only thing I remember is it has 512MB RAM and 40GB HDD.
That is not the problem. My box has 512 MB RAM and a 40 GB HD and I am running CentOS 5.2 (32 bit). It may be something else, such as the motherboard.
Le mardi 30 décembre 2008 à 19:37, MHR a écrit :
You need to supply more information about this "old" machine - that could be the problem....
Motherboard is EliteGroup P4VXASD+ Processor is Intel Celeron 4A 2GHz Chipset is VIA VT8753/A/E Apollo P4V266/A/E Videoadapter is nVIDIA RIVA TNT2 M64 (32MB)
Regards, Alain
On Tue, Dec 30, 2008 at 1:28 PM, Alain PORTAL alain.portal@free.fr wrote:
I'm trying to install Centos 5.2 on an "old" machine in graphical mode (to use LVM) but anaconda crash. I made several tries with various installation resquest (the last is "Server", "Customize after installation") unfortunately, it's reproductible:>
As Mark (mhr) mentioned, it might not run on that box. I suggest you download the "Live CD" for CentOS 5.2 and test it with that. If it doesn't run on the Live CD, it probably is not going to run on that box..... GL
Le mardi 30 décembre 2008, Lanny Marcus a écrit :
As Mark (mhr) mentioned, it might not run on that box. I suggest you download the "Live CD" for CentOS 5.2 and test it with that. If it doesn't run on the Live CD, it probably is not going to run on that
I didn't think about that. Thanks. I'll make a try tomorrow.
Regards, Alain
Le mardi 30 décembre 2008 à 19:53, Lanny Marcus a écrit :
As Mark (mhr) mentioned, it might not run on that box. I suggest you download the "Live CD" for CentOS 5.2 and test it with that. If it doesn't run on the Live CD, it probably is not going to run on that box..... GL
Trying in graphical mode and text mode, I get a lot of errors message such as: SQUASHFS error Unable to read page, block 2a673ccd, srclength 65536 SQUASHFS error zlib_fs returned unexpected result 0xfffffffd size 5230
I verified iso image (sha1sum) before burning it. k3b (Fedora Core 6) report burn successful. Is there a way with the Live CD to check the media before using it?
Regards, Alain
On Wed, 2008-12-31 at 13:36 +0100, Alain PORTAL wrote:
Le mardi 30 décembre 2008 à 19:53, Lanny Marcus a écrit :
As Mark (mhr) mentioned, it might not run on that box. I suggest you download the "Live CD" for CentOS 5.2 and test it with that. If it doesn't run on the Live CD, it probably is not going to run on that box..... GL
Trying in graphical mode and text mode, I get a lot of errors message such as: SQUASHFS error Unable to read page, block 2a673ccd, srclength 65536 SQUASHFS error zlib_fs returned unexpected result 0xfffffffd size 5230
I verified iso image (sha1sum) before burning it. k3b (Fedora Core 6) report burn successful. Is there a way with the Live CD to check the media before using it?
Just a thought along another tack: have you verified the hardware is good? That thouyght strikes me because of the nature of the messages you are getting. I have a mainboard with the VIA 8237 chipset and nvidia graphics (I don't recall which model ATM) and run 5.x w/o problems.
If the unit has been around a long time, try re-seating all the memory, cables, PS connectors, etc. Have you run a memtest86(?). Does a visual inspection show any signs of distress on the board (discoloration of capacitors, leaking, etc)?
Check BIOS settings too - maybe something set wrongly? Is the CMOS battery good? Do BIOS settings hold?
Regards, Alain
HTH
Le mercredi 31 décembre 2008 à 13:47, William L. Maltby a écrit :
Have you run a memtest86(?).
memtest86 found several errors during test #5. No more time today to test which memory is defective, I'll try next year ;-) which means next week.
Thanks!
Alain
Le mercredi 31 décembre 2008 à 14:22, Alain PORTAL a écrit :
Le mercredi 31 décembre 2008 à 13:47, William L. Maltby a écrit :
Have you run a memtest86(?).
memtest86 found several errors during test #5. No more time today to test which memory is defective, I'll try next year ;-) which means next week.
Finally, the two RAM were defective... I changed them with 2x512Mo KingStone and I successfully installed CentOS 5.2 ;-)
<voice mode="Droppy">I know what? I'm happy...</voice>
Regards, Alain
On Tue, 2009-01-06 at 19:00 +0100, Alain PORTAL wrote:
Le mercredi 31 décembre 2008 à 14:22, Alain PORTAL a écrit :
Le mercredi 31 décembre 2008 à 13:47, William L. Maltby a écrit :
Have you run a memtest86(?).
memtest86 found several errors during test #5. No more time today to test which memory is defective, I'll try next year ;-) which means next week.
Finally, the two RAM were defective... I changed them with 2x512Mo KingStone and I successfully installed CentOS 5.2 ;-)
<voice mode="Droppy">I know what? I'm happy...</voice>
I'm glad it worked out. I took the liberty of adding [SOLVED] to the subject line to help others that may have a similar problem in the future.
Regards, Alain
Le mardi 06 janvier 2009, William L. Maltby a écrit :
I'm glad it worked out. I took the liberty of adding [SOLVED] to the subject line to help others that may have a similar problem in the future.
Oups! Sorry, I should have to do it. But now, why (how) is it solved? Because a new good RAM, or because increased RAM quantity? ;-)
Regards, Alain
On Tue, 2009-01-06 at 20:40 +0100, Alain PORTAL wrote:
Le mardi 06 janvier 2009, William L. Maltby a écrit :
I'm glad it worked out. I took the liberty of adding [SOLVED] to the subject line to help others that may have a similar problem in the future.
Oups! Sorry, I should have to do it. But now, why (how) is it solved? Because a new good RAM, or because increased RAM quantity? ;-)
Who knows? Unless previous memory was _very_ small, it's most likely due to new good ram. But the important thing, I think, is that the memtest showed bad ram (IIRC, you said that) and you replaced the memory. So, the cause of the anaconda failure was seemingly bad memory.
If that cause and effect is true, the symptom was caused by the bad memory and that solved the problem. If someone has a problem with anaconda in the future, they will be reminded that install failures can be caused by memory problems, among other possibilities.
IIRC, graphic install needs >= 512MB and text installs need >= 256MB. But neither of those issues would cause a memtest failure. Poorly seated DRAM can cause it, memory _does_ age and can go bad and bad spots that were undetected in the factory QC process can also suddenly appear. Other things, like weak/failing PS can also cause symptoms.
Regards, Alain
<snip>
-
Alain PORTAL -
Alain PORTAL -
Lanny Marcus -
MHR -
William L. Maltby