| systemdlete | fsmithred, right you are. I meant hibernate/resume. But the point is, it isn't working. | 01:23 |
|---|---|---|
| rrq | don't you also have to declare resume device for initramfs-tools ? | 01:28 |
| rrq | used to be by editing /etc/initramfs-tools/conf.d/resume | 01:30 |
| rrq | maybe grub has subsumed that nowadays (it tends to expand) | 01:31 |
| systemdlete | you know, rrq, that's it! Thank you!!! | 02:12 |
| systemdlete | I cannot seem to locate dox that cover this. | 02:12 |
| systemdlete | for some reason, still not working. Now pm-hibernate works, as far as shutting down the system. But the boot hangs finally telling me it gave up waiting on resume device | 02:23 |
| systemdlete | I still think there is another piece of this. | 02:23 |
| systemdlete | (not sure why) | 02:24 |
| fsmithred | make sure uuid is correct | 02:38 |
| systemdlete | yup. Just did that. | 02:38 |
| systemdlete | uh... just erase erase erase. Put the wrong UUID in one of them... | 02:45 |
| systemdlete | pebcak | 02:49 |
| systemdlete | (totally) | 02:49 |
| systemdlete | it works | 02:49 |
| systemdlete | I am going to write myself a note this time... | 02:49 |
| systemdlete | only thing that doesn't quite make sense to me is why the resume device must be specified in 2 places, and yet I don't see how pm-utils will know where to save the system (assume there could be multiple swap devices) | 02:52 |
| systemdlete | or is it initramfs that tells pm-utils that, and why we update it? hmmm. | 02:53 |
| mason | systemdlete: My understanding is that you only need it in /etc/initramfs-tools/initramfs.conf | 03:06 |
| mason | The kernel handles hibernating and resuming. | 03:06 |
| mason | I could be confused as it's been a while since I've used that. | 03:06 |
| ibanja | ionos stopped installing screen on hosted servers. Can screen be compiled into a single binary? | 03:19 |
| ibanja | I want to upload it to my ~/bin directory. | 03:20 |
| systemdlete | mason, the ubuntu instructions seem to concur with your assessment also. They say that prior to a certain release, you had to do both. | 03:21 |
| systemdlete | anyway, my testing is not going so well. I am using daedalus, no login manager, invoking startxfce4 at login. The desktop comes up, then I sudo in order to pm-hibernate. the system shuts down. When it reboots, it does bring me back, but only to my shell that I launched xfce from. It does not restore the desktop, complaining that it lost connection to the x11 server | 03:23 |
| systemdlete | (forgot exactly what it said, but that is the gist) | 03:24 |
| mason | ibanja: You can install it locally, yes. You can install it into a local hierarchy in your home directory, and symlink it into ~/bin, or set your PATH accordingly. | 03:24 |
| systemdlete | Also, during the shutdown procedure in pm-hibernate, a console message appears on a blank screen by itself indicating that some gfx function (rings) has already been started, something like that. | 03:25 |
| mason | systemdlete: That sounds like it's not writing your RAM out to disk. | 03:25 |
| systemdlete | (and this worked so well when I did it months ago, not sure what happened) | 03:25 |
| mason | systemdlete: Or is it writing it out, and you're getting the shell you expect, but not the X session? | 03:25 |
| systemdlete | but it does restore the system, just not the desktop. | 03:25 |
| mason | What graphics card? Maybe there's some special handling required. | 03:26 |
| systemdlete | I don't expect the shell, no. I expect the desktop I launched from that shell. | 03:26 |
| systemdlete | radeon | 03:26 |
| systemdlete | (which one?) | 03:26 |
| systemdlete | I am running mostly AMD hardware here. | 03:26 |
| systemdlete | just fyi | 03:26 |
| systemdlete | I think I have one Intel video card | 03:26 |
| systemdlete | but it is not in use atm | 03:27 |
| mason | I can't remember the last time I tried it with anything other than Intel. | 03:27 |
| ibanja | mason: Is that by installing a deb package? | 03:29 |
| systemdlete | Maybe I was working on chimaera host at the time, idk. I'll try that and see if my luck changes | 03:29 |
| mason | ibanja: No, you'd grab the source, ./configure ; make ; make install PREFIX=/home/ibanja/localbins or something similar | 03:29 |
| mason | Half a sec, that's more tuned towards using Stow or something. ./configure --prefix=/home/ibanja/localbins | 03:30 |
| mason | (Setting PREFIX as an environment variable builds the binaries to think everything's in /usr/local/ and you use Stow to set up a symlink farm accordingly, but you want something simpler that *knows* it's in /home/ibanja/somewhere) | 03:31 |
| mason | ibanja: Anyway, simple way to get the source: apt source screen | 03:31 |
| ibanja | I will try that. I tried the git source and it complained there was no Make file. | 03:33 |
| mason | ibanja: ./configure creates your Makefile | 03:34 |
| mason | ibanja: If you've already got the source from git, you can use that just as I've described. | 03:34 |
| ibanja | make is complaining; make: *** No targets specified and no makefile found. Stop. | 03:38 |
| ibanja | I ran .configure --prefix=/home/ibanja/localbins | 03:39 |
| ibanja | using my path | 03:39 |
| mason | Did that not return an error? | 03:39 |
| mason | ./configure from inside thre source directory. | 03:39 |
| ibanja | no error and run from inside src fir | 03:40 |
| ibanja | dir | 03:40 |
| mason | ls -l and make sure there's a ./configure script in your directory. If not there's an additional step. Once you've run it, verify that you have a Makefile there | 03:40 |
| ibanja | there is a ./configure file. It seems to run and do a bunch of checking. | 03:41 |
| mason | There should be a Makefile after it's done. | 03:41 |
| ibanja | excuse me. the autogen did the checking | 03:41 |
| ibanja | what is the step for no configure file? I will try it in the apt source download instead of git sources. | 03:42 |
| ibanja | there is no configure in the apt source sources. | 03:43 |
| mason | Oh, hm, just occurred to me that you might also need to have some local build dependencies. Checking. | 03:43 |
| ibanja | I assume ./configure.ac is something different? | 03:44 |
| mason | ibanja: Run 'autoconf' and perhaps it'll turn your configure.ac into a configure script. | 03:45 |
| mason | Since you're building outside of the Debian tooling, build-deps might be less obvious. Your ./configure script ought to tell you what it needs when you run it. | 03:47 |
| ibanja | it is saying autom4te: error: cannot create autom4te.cache in home/ibanja/src/screen/screen-4.9.0: Permission denied | 03:49 |
| ibanja | do I need to be root? | 03:49 |
| ibanja | hold... i ran as root and got a lot of output | 03:49 |
| mason | You should only need to own the directories in which it's trying to write. You don't want to be root. | 03:49 |
| mason | That said, if you have root on the box, why not just install the screen package? | 03:50 |
| ibanja | I don't have root on the server. | 03:51 |
| mason | "hold... i ran as root and got a lot of output" ? | 03:51 |
| ibanja | that's why I wanted to compile to a single binary. | 03:51 |
| mason | Are you wanting to build locally and then shlep it over to some third-party server then? | 03:52 |
| mason | You can always take your resulting /home/ibanja/localbins and move that over. | 03:52 |
| ibanja | I want to build screen in my home directory and then upload to my server. | 03:52 |
| Xenguy | (There's always tmux if that's at all relevant) | 03:52 |
| mason | Xenguy: Oh, right, there's that too. Worth seeing if that's there. | 03:53 |
| ibanja | I could try dpkg-deb -x screen_4.9.0-4_amd64.deb /home/pathtobin | 03:53 |
| ibanja | what is tmux? | 03:53 |
| Xenguy | NG screen-like program | 03:54 |
| mason | It's like screen(1), just a different program doing mostly the same thing. | 03:54 |
| mason | It doesn't do the serial console stuff screen(1) does but is otherwise pretty similar. I don't use it myself, mostly because I'm used to screen(1) and screen(1) hadn't broken recently. | 03:54 |
| Xenguy | I even found out that screen is still being maintained | 03:55 |
| Xenguy | I use both, just in case | 03:55 |
| ibanja | I'm trying to keep it small because ionos gave me a lifetime price for unlimited space at a lifetime price never to be increased. However, after the fact they limited the number of inodes I can have and after a number of years increased the price and dropped screen. | 03:58 |
| ibanja | that was a bit of a run-on sentence... | 03:59 |
| ibanja | I'm going to try the dpkg-deb method and strip out the unnecessary bits. Thanks for the help. | 04:00 |
| mason | Hm. Busybox doesn't seem to include anything like that. | 04:01 |
| ibanja | that's an idea... checking. | 04:03 |
| mason | ibanja: It's worth making sure that you let processes linger, that they won't kill them on you. | 04:14 |
| ibanja | The main reason I want screen is because my ISP is so lousy I keep getting dropped. It's frustrating. So if I get dropped I can log back in and reconnect to screen. | 04:17 |
| ibanja | I don't think I'll have to many lingering processes. they probably would kill them too. | 04:17 |
| mason | screen itself would be such a lingering process | 04:18 |
| ibanja | I don't see a busybox screen | 04:18 |
| mason | Right. It doesn't include it. I checked. | 04:18 |
| ibanja | I am usually not on for too long. The problem is that sometimes my ISP drops the connection many times in a day. The last week I probably have been disconnected from the internet about 15-50 times a day. They just fixed it today. | 04:20 |
| ibanja | I try to do any file changing on my home computer then rsync up to the server, but sometimes I have to be on the server. | 04:21 |
| systemdlete | on boot, I get a (harmless) prompt: diskfilter writes not supported. I know exactly why, too. My boot drive is on a raid, and I have (tried to) set GRUB_DEFAULT=saved along with GRUB_SAVEDEFAULT=true | 04:22 |
| ibanja | mason: I'm going to log off... thanks again for the help. | 04:22 |
| systemdlete | so it is trying to write to a disk to update the grub selection, which it can't do, due to the lack of raid support in grub. | 04:22 |
| systemdlete | this is noted as a bug by ubuntu and some others, but it has been many years and it still fails and a bug fix (if there really is one) is not forthcoming it seems | 04:23 |
| systemdlete | (anyone reading this may file this as BELLYACHE; WONTFIX) | 04:23 |
| systemdlete | I suppose I could arrange to include the raid and lvm software in grub, but I wonder if that alone will solve it | 04:24 |
| systemdlete | Or maybe EFI? On this particular machine, the mb does not have EFI, so I would need to use an alternate, but that might be simpler than trying to do battle with grub shortcomings? | 04:28 |
| * systemdlete wanders back to The Curse of Oak Island (they're getting ready to break trhough the bottom of the garden shaft...) | 04:29 | |
| mason | systemdlete: I'll chop out some time to do an install with hibernation to verify it works for me, and I'll describe what I do. | 05:25 |
| mason | Did you say this was on Daedalus? | 05:26 |
| systemdlete | mason, I have it working. My complaint, above, was not really related. | 07:38 |
| systemdlete | The reason it was not working was because I had mismatched the partition uuid. | 07:38 |
| mason | Oh, I saw that and missed it somehow. Good then. | 07:53 |
| systemdlete | gnarface, rrq: I upgraded my apt-cacher-ng to the backports version and, since doing so, I have not seen the connection problems or errors derived from those. | 19:25 |
| systemdlete | The system hosting apt-cacher-ng is actually chimaera, so it is one major cacher version back by default. The upgrade brings it up to deadalus. | 19:26 |
| systemdlete | I need to upgrade that system at some point... | 19:26 |
| systemdlete | so it appears that the problem is not with the servers or rr at this early point; then again, I haven't had any real updates since then. I have, however, done some installs since then, and without problems. | 19:27 |
| systemdlete | I am not 100% sure this is resolved with the cacher upgrade, but I am hopeful. | 19:28 |
| systemdlete | s/major/minor/ | 19:28 |
| systemdlete | 3.7.4 vs 3.6.4 | 19:29 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!