[3387] in SIPB_Linux_Development
outstanding bugs
daemon@ATHENA.MIT.EDU (Jonathon Weiss)
Mon Mar 19 02:40:52 2001
Date: Mon, 19 Mar 2001 02:40:29 -0500 (EST)
Message-Id: <200103190740.CAA06282@Bearing-An-Hourglass.mit.edu>
To: linux-dev@MIT.EDU
From: Jonathon Weiss <jweiss@MIT.EDU>
I suggested that someone needed to put together a list of unresolved
things that needed to be dealt with before the install went
production. No one volunteered, but at least one person whined at me
for not doing it myself (despite the fact that I have a lot of other
stuff to do). Anyhow, here it is, no more whining:
In general I haven't included anythign that that seems like a RedHat
problem (including a number of the network card issues). numbers in
brackets are transaction numbers in linux-dev. These are in 3
categories: showstoppers, bugs that aren't showstoppers, and
mysterious lossage. We need to empty the showstopper list before
going to production (and I will allow for the possibility that some of
the things I classified as showstoppers may be reclassified to a lower
level.)
showstoppers, IMHO:
[2950] [3062] [3323] [3361] [3368]
resolv.conf has nameserver supplied during install, not localhost.
fixed (has this been tested)
[3213] [3383] [3384]
install lists bogus times in /etc/athena/version
believed fixed, needs to be tested and propogated to charon
[2955]
Somehow, my kerberos realm defaulted to "EXAMPLE.COM".
This prevents athena users from logging in with xlogin.
I'm not sure who owns the /etc/krb5.conf file:
bash# rpm -qf /etc/krb5.conf
athena-krb5-8.4-10
krb5-configs-1.1.1-9
We need to figure out if krb-configs is sometimes geting installed
after athena-krb5 for some reason. There may have been discussion of
doing an install and saving the install log and seeing whether there
was anythign interesting in it, but no one ever dealt. Does this need
to be requalified as mysterious lossage since it is working for other
people?
[2989]
Currently we install a null release RPMs file. This works except that
if the user selected not to install some Athena package as part of a
custom install it will be installed at next update. Alternatively if
the release drops some package it will get left on the system.
The release may drop RPMs in patch releases for security reasons
(this happened to gnorpm between 8.4.15 an 8.4.17).
[3044] [3056] [3109] [3110] [3112] [3113] [3114]
We run getty on other vt's. I've seen a lot of back and forth on
this, as to whether we decided to leave it or fix it to be like the IS
install. At the very least we need to get our story straight. If we
leave it I think we need to make sure the customization gets into the
private-workstation-owners-guide.
[3048] [3051] [3052] [3055] [3058] [3059]
AFSCLIENT AFSADJUST AFSSRV
A couple of related issues here: None of these variables seems to be
getting installed in rc.conf by the expected mechanism. The frist two
are being added by the post-install script. At the very least we need
to have post-install deal with AFSSRV too, and have the post-instal
only add the variables if they aren't already there (right now it adds
them without thinking.) Better, of course would be to find and fix
the actual problem. There may have been discussion of saving the
install log form an install and seeing if it contained anythign
interesting (like the indication of athena-afs gettign installed
before athena-afs or something.)
[3073]
User with custom install problmes. Did we agree to and implement the
removal of the custom install option?
[3316] [3317] 3318]
fix pcitable on sipb-nfs (not really a show stopper, but a trivial
file propogation from the floppy to charon as I understand it)
[3354]
We're installing old versions of some RedHat packages. (I thought
we'd fixed this, did I just miss the mail in my scan?)
bugs that should get fixed, but probably aren't show stoppers
[3217]
lilo auto-installed on the MBR (as opposed to asking the user where to
install)
[3225] [3226]
mail to register_ws should always have FQDN
[3236] [3228]
mkserv and version of 8.4.0 (mkserv reports errors if you run it
before you update from 8.4.0 to the current patch release)
[3297]
If you use dhcp to get your address during the install the wrong thing
(nothing?) gets put in rc.conf
Mysterious lossage that was never formally resolved, but probably
isn't a showstopper:
[3088]
weird delays using the graphical installer
[3187] [3188] [3189] [3190] [3191]
sly: 4g partition ended up with a 1g filesystem on it
[3212]
hangs installing tetex
[3339]
hang in post-install
[3362]
bunzip2 error
[3381]
bugs using back button
