6.06->6.10 hangs on both i386 and amd64, log filled with "too many values to unpack" (ValueError)

[Timo Jyrinki]

i386 dapper, quite clean install with some packages from universe and about three from 3rd party source which is correctly detected and disabled by the update-manager. another machine with anoter amd64 dapper install, no packages outside main and universe.

/var/log/dist-upgrade/main.log gets filled with the following very soon after the Big debugs about packages to remove and install:
2006-10-26 18:57:43,394 DEBUG Free space on /: 15167369216
2006-10-26 18:57:43,395 DEBUG Dir /usr mounted on /
2006-10-26 18:57:43,395 DEBUG Dir /var mounted on /
2006-10-26 18:57:43,395 DEBUG Free space on /boot: 250615808
2006-10-26 18:57:43,781 DEBUG required download: 693437142.0
2006-10-26 18:57:43,782 DEBUG free on /var/cache/apt/archives/: 15167369216
2006-10-26 18:57:44,122 DEBUG need additional space: 318287872.0
2006-10-26 18:57:44,456 DEBUG /usr on same fs as /var/cache/apt/archives/, taking dl-size into account, new free: 14473932074.0
2006-10-26 18:57:44,457 DEBUG using safety buffer: 104857600
2006-10-26 19:20:33,041 ERROR got ValueError from InstallPrgoress.updateInterface. Line was 'pmstatus:libc6:0.245512:libc6:n asetusten s<C3><A4><C3><A4>t<C3>
<B6>
' (too many values to unpack)
2006-10-26 19:20:33,080 ERROR got ValueError from InstallPrgoress.updateInterface. Line was 'pmstatus:libc6:0.245512:libc6:n asetusten s<C3><A4><C3><A4>t<C3>
<B6>
' (too many values to unpack)

(and continuing with the four InstallPrgoress lines until infinity)

"asetusten säätö" just means "configuring"
Currently 19MB log on the other machine and 13MB on the machine. I guess I have to kill the update-manager processes and try something else.

[Timo Jyrinki (timo-jyrinki)]

term.log for the i386 machine. Amd64 machine's term.log also stops somewhere, but in a different place. There the last line is "Removing libtasn1-2-dev"

[Timo Jyrinki (timo-jyrinki)]

Ends similarly on the amd64 machine. And the locale used on both machines is fi_FI.UTF-8.

[Michael Vogt (mvo)]

It looks like this is locale releated. I have got a similar report ( #68553) from a finish user as well.