also sprach Bas Wijnen <[EMAIL PROTECTED]> [2006.08.02.1051 +0100]: > > The loop should stay endless, IMHO, to ensure that the user passes > > a proper device string (or 'all' or 'none'). > > The loop being endless if the user actually gets to answer the quesion is > fine. But my debconf priority is higher than "low", so the question isn't > asked at all, it just loops until I hit ctrl-C. > > But you did just point out a workaround. :-) I can just set my debconf > priority to low and fill in "none" or "all".
ouch, this is not good. However, I wonder if it's fixed now. My analysis is that the system autodetected /dev/md_d0 but then didn't accept that in the loop and didn't give you a way to change that. Now it should be able to deal with /dev/md_d0, so even at priority low, things should be fine, right? > > but undoubtedly, the problem you point out should be fixed. > > That too, yes. :-) But I think the loop should also not be > endless if the user isn't asked anything. If the user isn't asked anything, the question should have been auto-answered in a way not to enter an endless loop. This is what happened previously, but it should not happen now anymore, right? > Well, that'd patch the current version, not the one yet to be > unpacked and installed. The problem is that the install doesn't > succeed, and thus the old package is still the one which has its > config file there. I'd need to rebuild the package, which is > a bit too much trouble at the moment. You can dpkg --unpack /var/cache/apt/archives/mdadm_2.5.2-9_*.deb and then apply the patch, then dpkg --configure -a Does this work? > > - /dev/md[0-9]*|/dev/md_d[0-9]*) > > + /dev/md[0-9]*|/dev/md_d[0-9]*|/dev/md/[0-9]*|/dev/md/d[0-9]*) > > Also, while this would of course solve the "normal" configurations, the > problem would still exist for people who have told udev that their device is > called /dev/md_disk_0 or something. That may be really rare, but it shouldn't > lead to a hanging install process. ;-) Mh, I agree. So basically I have to loosen the sanity checks... > > and finally try to reboot > > I'd prefer not to, I'm quite happy with my >1 year uptime. :-) > > Of course if there's a good reason to reboot, I'd do it, but > AFAICS things still work pretty well. I think any change to the system is a good reason to reboot, or else you may find a power outage to cycle the system and expose an error at a time when you don't have the time or nerves to fix it. -- Please do not send copies of list mail to me; I read the list! .''`. martin f. krafft <[EMAIL PROTECTED]> : :' : proud Debian developer and author: http://debiansystem.info `. `'` `- Debian - when you have better things to do than fixing a system
signature.asc
Description: Digital signature (GPG/PGP)
