Using rescuecd !?
Hi all;
Let me start by saying I am NOT in immediate need of using the rescuecd. But, I have been on other occasions. I have been unable to find a manual or guide on using rescuecd. I have found HOWTO's that give me the exact button pushes or command line instructions to use and managed to fix the problem at the time. I have not, however, found a text that explains what I am doing or why. I have a list of about 20 questions that could be answered and explained so that I would feel confident in using the rescuecd rather than feeling like I am making it up as I go along. There are two small issues that I would like to test here to see if there is agreement: 1) Could the word 'sysimage' for root's home file system be changed to something else more meaningful? To me, when I first had to use rescuecd, 'sysimage' was one extra bit of jargon I had to learn to translate when I was already in an agitated state because my system was dead. 2) Could the prompt be set up to show me more information than sh-3.2#? It would make it easier for a new user to keep track of whether they were working in [rescuecd/sh-3.2]# file system or the [CASE/sh-3.2]# file system (CASE is the name of my host machine), if the prompt was more explicit. Remember when someone, particularly a new user, is in rescuecd for the first couple of times, they are usually in an emotional state characterized by frustration, fluster and fear that they have just blown their entire system. Adding a burden of unnecessary, incomprehensibility is just a disservice. Any ideas? Can these changes be considered a bug? -- Regards Bill -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list |
Using rescuecd !?
William Case wrote:
Hi all; Let me start by saying I am NOT in immediate need of using the rescuecd. But, I have been on other occasions. I have been unable to find a manual or guide on using rescuecd. I have found HOWTO's that give me the exact button pushes or command line instructions to use and managed to fix the problem at the time. I have not, however, found a text that explains what I am doing or why. I have a list of about 20 questions that could be answered and explained so that I would feel confident in using the rescuecd rather than feeling like I am making it up as I go along. There are two small issues that I would like to test here to see if there is agreement: 1) Could the word 'sysimage' for root's home file system be changed to something else more meaningful? To me, when I first had to use rescuecd, 'sysimage' was one extra bit of jargon I had to learn to translate when I was already in an agitated state because my system was dead. This is used for consistency with the installer. To the installer, what's in that directory is the image of what will eventually be the installed system. 2) Could the prompt be set up to show me more information than sh-3.2#? It would make it easier for a new user to keep track of whether they were working in [rescuecd/sh-3.2]# file system or the [CASE/sh-3.2]# file system (CASE is the name of my host machine), if the prompt was more explicit. This is the base, unconfigured prompt, as used in single-user mode, the second virtual terminal of the installer, and as a user when logging in with no home directory mounted. It indicates that no user configuration could be read from disk. Remember when someone, particularly a new user, is in rescuecd for the first couple of times, they are usually in an emotional state characterized by frustration, fluster and fear that they have just blown their entire system. Adding a burden of unnecessary, incomprehensibility is just a disservice. This isn't unnecessary incomprehensibility, it's failsafe simplicity. Any ideas? Can these changes be considered a bug? The existing behavior is simple, functioning as designed, widely documented, and familiar to a lot of people. There's nothing inherently wrong with rescue mode *for what it was designed to do*. The problem is that rescue mode wasn't designed for a new user. We really need a user-friendly recovery console, but it should be an application that works on top of the existing rescue mode that the experts already know, not a replacement for it. -- Chris -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list |
Using rescuecd !?
Chris Snook wrote:
William Case wrote: ... I have a list of about 20 questions that could be answered and explained so that I would feel confident in using the rescuecd rather than feeling like I am making it up as I go along. ... This isn't unnecessary incomprehensibility, it's failsafe simplicity. Any ideas? Can these changes be considered a bug? The existing behavior is simple, functioning as designed, widely documented, and familiar to a lot of people. There's nothing inherently wrong with rescue mode *for what it was designed to do*. The problem is that rescue mode wasn't designed for a new user. We really need a user-friendly recovery console, but it should be an application that works on top of the existing rescue mode that the experts already know, not a replacement for it. It would be nice to prompt user yes/no to chroot to the users system if found, rather than requiring the typing. Other things that I consider could be useful; perhaps in a ncurses text interface like setup: - test if there is bootable grub -> suggest needs fixing and perform best guess grub-install xxx - test if there is grub.conf points to kernel bits, and root= is OK. - test if there is at least a kernel's boot bits installed / available - test if the /etc/fstab points to legit things {eg for times when eg a partition has become unmountable, or swap part is no longer there}. - offer to backup and mod fstab to # non-essential bits {like non-root/boot drives} - test rpm command works - test yum works - offer to rpm -V important == needed to boot packages, mentioning missing packages considered essential. {eg after --nodeps fu's} - offer to rpm -Va - offer to fsck - offer to selinux relabel {listing files modified, and logging for later perusal}. - offer to make some room on an empty fs - offer to install a kernel from release|updates|updates-testing - offer info on how to show the grub menu - and set grub.conf to wait for user - offer to copy a heap of kernel boot parameters into a grub entry - so that the user can delete most and try one at a time out. - {better} offer to copy a current grub entry, name it X modified, and allow the user to select only compatible parameters to try. {and set it default} - offer to copy a current grub entry and append runlevel - with descriptions - offer to install rescue mode onto the hard disk in a separate partition or in the /boot if there is enough room, and add a grub option for it. - an option to not chroot, and instead attempt to start the user's system as a virtual machine ...? - offer to remove options it added to grub/ fstab etc. - change rescue cd to have a basic menu with rescue as the default option {rather than an attempt at upgrade / install} - run package-cleanup --problems - check for issues like the F6 wrong arch kernel install. - update clamAV defs and perform a system scan. - update chkrootkit / rkhunter and run a rk scan. - gui to run gparted or s-c-lvm - lvm tools assistance. - test root/ a user login files {passwd etc} are coherent - all tests, from the lowest level - boot up, or a particular test if the user has an idea. - easy way to ftp/http a file from a local network {eg no internet access} or from the internet {eg an older kernel or whatever} ~ mc ? Well, that is my twenty or so things that I have had to do over the last few years. William, what would you add to that ? By the way, you could create a fedoraproject wiki page so that interested parties could put/edit their ideas into one place. Once this becomes considered you could file a RFE [request for enhancement] in bugzilla. DaveT. -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list |
Using rescuecd !?
Hi David;
On Thu, 2007-11-29 at 23:36 +1100, David Timms wrote: > Chris Snook wrote: > > William Case wrote: [snip] > Well, that is my twenty or so things that I have had to do over the last > few years. William, what would you add to that ? > > By the way, you could create a fedoraproject wiki page so that > interested parties could put/edit their ideas into one place. Once this > becomes considered you could file a RFE [request for enhancement] in > bugzilla. If we are going to live in a fantasy land of hope, I would add just one thing - offer to open in an editor /etc/X11/xorg.conf. I can figure my way around resecuecd now. But I remember vividly the first couple of times I had to use it. An explanatory manual would have been helpful. Perhaps as a downloadable PDF on the Fedora site/wiki that I could have printed out or read from another computer or OS. As well, some onscreen indication of which file system I had mounted and was working in. I kind of agree with Chris Snook's sentiment -- but everyone has to use rescuecd for the first time -- at least once. -- Regards Bill -- fedora-list mailing list fedora-list@redhat.com To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list |
| All times are GMT. The time now is 05:45 AM. |
VBulletin, Copyright ©2000 - 2013, Jelsoft Enterprises Ltd.
Content Relevant URLs by vBSEO ©2007, Crawlability, Inc.