Re: [gentoo-user] Re: Gentoo is supporting officially Snap packages?
On Sat, Jun 18, 2016 at 4:01 AM, Mickwrote: > On Thursday 16 Jun 2016 21:25:01 J. García wrote: > > How does Nix compare to flatpack, docker, snap, et al. from a gentoo > perspective? > Nix is a similar sort of approach. I don't think they run apps in containers (though they probably could if they wanted to do the work and a lot of bind mounting). The reality is that they're a form of bundling, but the bundled libs can be shared. Basically everything is linked to uniquely identified dependencies. So, a package isn't just linked to zlib, or even a particular version of zlib, but a particular build of zlib. However, 15 different packages could all depend on that same build. So, you potentially don't get the same kind of memory duplication that you do with outright bundling. However, if you install a new version of zlib on your system, nothing will actually use it, unless those packages are themselves updated. So, in that respect it is just like bundling. Since the libraries you're running with on your box are exact copies of the binaries the packager was using, you're going to get the same experience the packager did when they were testing their package. So, that's the big upside. There are no conflicts or collisions either, since every package is installed in what amounts to a private namespace. You can have 14 different packaged builds of zlib-1.2.3 if you want to, with different builds being used by different applications. This is just my understanding based on having looked into NixOS a bit out of curiosity. Somebody closer to the project should feel free to correct any errors I made. There are obviously pros and cons to this approach. -- Rich
Re: [gentoo-user] Recommended way to shut down akonadi
On Saturday 18 Jun 2016 19:42:05 J. Roeleveld wrote: > On Saturday, June 18, 2016 05:45:31 PM Mick wrote: > > Thank you this did not work. I didn't know that /etc/local.d/ scripts run > > at shutdown, but in any case even if they do I suspect they don't run > > until X exits first. > > add the following to /etc/rc.conf: > > rc_local_need="postgresql-9.5" > (or whichever version you are currently using) > > That will force "local" to be stopped before "xdm" > > but as far as I am aware, the "local" service is normally started last and > stopped first. > As long as it's stopped before postgresql, you should be fine. Is don't fully understand how this may help with the problem I am experiencing? Namely, akonadi will *not* stop and shutdown its agents, no matter what scripts I've provided either in ~/.xsession or in /etc/local.d/killakonadi.stop, in order to release gracefully any connections to postgresql? Regardless, I added the stanza in rc.conf and still have the same problem. TBH I have a new problem too, which is that net.enp11s0 is now counting down and waiting for ...40, ...30, ...20 seconds and so on at boot up. Akonadi's wonderful agents won't give up at shut down. Thank you for your kind help. I think I'll give up for now and just use a terminal to stop akonadi before I shut down the PC. -- Regards, Mick signature.asc Description: This is a digitally signed message part.
[gentoo-user] Re: Virtual machines, and creating bootable images
Jigme Datse Yli-RAsku datsemultimedia.com> writes: > Anyone have some good starting points to think about this? Here is a link to some amd64 stage-4 images:: http://rsync5.us.gentoo.org/gentoo/releases/amd64/autobuilds/current-install-amd64-minimal/ Here is the Google Summer of Code proposal, where the devs would be willing to work with anyone who wanted to develop an updated method to stag3-4 installs. Oh, just in case you do not know, lots of folks like to make stage-4 images of servers or systems, for an easy backup/restore snapshop of critcal systems:: https://wiki.gentoo.org/wiki/Google_Summer_of_Code/2016/Ideas/ Stage4_Console_Configurator hth, James
Re: [gentoo-user] Recommended way to shut down akonadi
On Saturday, June 18, 2016 05:45:31 PM Mick wrote: > On Tuesday 14 Jun 2016 07:03:25 J. Roeleveld wrote: > > On Friday, June 10, 2016 10:32:25 PM Mick wrote: > > > On Friday 10 Jun 2016 19:54:44 J. Roeleveld wrote: > > > > Kde and I would assume Gnome as well have a method of running a script > > > > when > > > > starting and stopping a login session. > > > > > > > > This is usually used for ssh-agent and pgp-agent. > > > > > > Yes, KDE uses /etc/plasma/shutdown/10-agent-shutdown.sh and I guess > > > gnome > > > would have its equivalent, but I think this is only for any daemonised > > > services running on the desktop. > > > > > > I suspect that: > > > > > > /usr/bin/kdeinit4_shutdown > > > /usr/bin/kdeinit5_shutdown > > > > > > are used to shutdown gracefully any KDE apps. > > > > > > > You could try looking for something similar with the desktop/window > > > > manager > > > > of your choice. > > > > > > I tried the above shutdown commands but they didn't work. Unfortunately > > > enlightenment does not have anything available to stop desktop > > > applications > > > at shutdown. The dev's advice was to use .xinitrc or equivalent. > > > > > > > I think .xsession is run only during start and will not 'pause' during > > > > the > > > > session. > > > > > > Well my confusion is that my .xsession *was* working fine until a couple > > > of > > > weeks ago ... and all still works as expected when I run akonadictl stop > > > in > > > a terminal before I shut down. I don't know why the same command > > > behaves > > > differently in .xsession now. :-/ > > > > > > PS. When I just log out there is no delay. The problem only arises when > > > I > > > shutdown. I wonder if this is something to do with this darn sddm > > > display > > > manager ... > > > > There is a 90 second delay with PostgreSQL if there are still clients > > connected to the database. > > This should also be displayed on the screen if you see the shutdown-text. > > Yes, postgresql waits 90 seconds until all these akonadi attempts to > reconnect its agents finish, then X exits, postgresql shutsdown and then > system shuts down as expected. > > > I don't see how sddm can cause an extra delay. > > I can't either, but I am clutching at straws here. > > > You could add the commands to: > > /etc/local.d/killakonadi.stop > > > > (These are run as root, so you might need to do something like: > > /bin/su -s /bin/sh -c " " > > instead of the commands you have been putting in the scripts. > > Thank you this did not work. I didn't know that /etc/local.d/ scripts run > at shutdown, but in any case even if they do I suspect they don't run until > X exits first. add the following to /etc/rc.conf: rc_local_need="postgresql-9.5" (or whichever version you are currently using) That will force "local" to be stopped before "xdm" but as far as I am aware, the "local" service is normally started last and stopped first. As long as it's stopped before postgresql, you should be fine. > > I think the local-service is stopped before postgresql is stopped, so this > > might prevent the delay you are seeing? > > > > -- > > Joost > > I usually have a terminal open (more often than I have kmail) so I will need > to remember to run 'akonadictl stop' in the terminal before I shutdown the > machine. It's not ideal, but unless something can stop effectively akonadi > I can't exit X. -- Joost
Re: [gentoo-user] Recommended way to shut down akonadi
On Tuesday 14 Jun 2016 07:03:25 J. Roeleveld wrote: > On Friday, June 10, 2016 10:32:25 PM Mick wrote: > > On Friday 10 Jun 2016 19:54:44 J. Roeleveld wrote: > > > Kde and I would assume Gnome as well have a method of running a script > > > when > > > starting and stopping a login session. > > > > > > This is usually used for ssh-agent and pgp-agent. > > > > Yes, KDE uses /etc/plasma/shutdown/10-agent-shutdown.sh and I guess gnome > > would have its equivalent, but I think this is only for any daemonised > > services running on the desktop. > > > > I suspect that: > > > > /usr/bin/kdeinit4_shutdown > > /usr/bin/kdeinit5_shutdown > > > > are used to shutdown gracefully any KDE apps. > > > > > You could try looking for something similar with the desktop/window > > > manager > > > of your choice. > > > > I tried the above shutdown commands but they didn't work. Unfortunately > > enlightenment does not have anything available to stop desktop > > applications > > at shutdown. The dev's advice was to use .xinitrc or equivalent. > > > > > I think .xsession is run only during start and will not 'pause' during > > > the > > > session. > > > > Well my confusion is that my .xsession *was* working fine until a couple > > of > > weeks ago ... and all still works as expected when I run akonadictl stop > > in > > a terminal before I shut down. I don't know why the same command behaves > > differently in .xsession now. :-/ > > > > PS. When I just log out there is no delay. The problem only arises when I > > shutdown. I wonder if this is something to do with this darn sddm display > > manager ... > > There is a 90 second delay with PostgreSQL if there are still clients > connected to the database. > This should also be displayed on the screen if you see the shutdown-text. Yes, postgresql waits 90 seconds until all these akonadi attempts to reconnect its agents finish, then X exits, postgresql shutsdown and then system shuts down as expected. > I don't see how sddm can cause an extra delay. I can't either, but I am clutching at straws here. > You could add the commands to: > /etc/local.d/killakonadi.stop > > (These are run as root, so you might need to do something like: > /bin/su -s /bin/sh -c " " > instead of the commands you have been putting in the scripts. Thank you this did not work. I didn't know that /etc/local.d/ scripts run at shutdown, but in any case even if they do I suspect they don't run until X exits first. > I think the local-service is stopped before postgresql is stopped, so this > might prevent the delay you are seeing? > > -- > Joost I usually have a terminal open (more often than I have kmail) so I will need to remember to run 'akonadictl stop' in the terminal before I shutdown the machine. It's not ideal, but unless something can stop effectively akonadi I can't exit X. -- Regards, Mick signature.asc Description: This is a digitally signed message part.
[gentoo-user] Re: Virtual machines, and creating bootable images
Jigme Datse Yli-RAsku datsemultimedia.com> writes: > I am looking at creating a gentoo install which I can use for my hosted > servers. The documentation I have seen about doing so is to mount an > "ISO" and then dd the mounted device to a data drive, make the data > drive bootable (set it in the settings for hosting provider) and attach > it to a server. Testing:: http://user-mode-linux.sourceforge.net/ VM/container images:: There are many around. Rich0 put up one of my favorites, which included btrfs:: https://docs.google.com/document/ d/1VJlJyYLTZScta9a81xgKOIBjYsG3_VfxxmUSxG23Uxg/edit?pli=1 And if you are into openstack cluster images:: https://blog.mthode.org/posts/2016/Jun/gentoo-openstack-and-osic/ > I want to be able to build these images and test them in a virtual > machine on my local server, then end up doing what I need on the hosting > provider to get them in place. Yes we all want this. Truth is, it is very easy to make a misstep. Catalyst was the traditional tool for this this, but it is still usable (at least some claim so). ymmv. Some of the older scripts are very instructive for the steps, but the details will need to modified to fit your current needs. https://wiki.gentoo.org/wiki/Catalyst GRS:: https://wiki.gentoo.org/wiki/Project:RelEng_GRS Stage 4:: https://github.com/TheChymera/mkstage4 https://forums.gentoo.org/viewtopic-t-312817.html http://www.tutorials.chymera.eu/blog/2014/05/18/ mkstage4-stage4-tarballs-made-easy/ https://forums.gentoo.org/viewtopic.php?p=1751698#1751698 Random links:: http://opensource.dyc.edu/tinhat-howtocook https://forums.gentoo.org/viewtopic-t-995118.html http://www.syslinux.org/wiki/index.php?title=Isohybrid https://tails.boum.org/blueprint/bootstrapping/installer/ https://github.com/mikemol/gentoo-install/blob/master/gentoo-install.sh NEW pathways to robust, automated installation:: https://github.com/gentoo/stager/blob/master/road_map.md https://calamares.io/ app-admin/calamares Ansible:: http://blog.jameskyle.org/2014/08/automated-stage3-gentoo-install-using-ansible/ https://github.com/gentoo-ansible > Quick research hasn't given me any clear guidance as to how I can export > the virtual machine (ie. what even can do this) to an image that can be > copied to a drive to produce a bootable drive. > Anyone have some good starting points to think about this? Once you have worked through the gentoo handbook (maybe a few times) all of the methods do start to make more sense. I have been slowly been gathering and testing up old and new methods to automate gentoo installs, regardless of platform:: on discrete hardware, vm, container, hpc or embedded. All have similarities and mostly differ in how you set up the storage (ram, usb, disk, flash, etc etc). > Jigme Datse Yli-Rasku This list is not exhaustive. Before folks will help you very much, make sure you have gone through at least a few gentoo handbook installs and know those steps well. Do post back, as auto/image installs are still a pita for me, as it does take time to smooth out the details on any installation method for gentoo I've been in pursuit of the holy grail of gentoo install. A custom hardware rack/apparatus, where I can plug in anything and perform a myriad of gentoo installs, all with just a few click. ymmv. I have most of the pieces, but ADHD stills has me blocked GOOD HUNTING! hth, James
Re: [gentoo-user] Re: Gentoo is supporting officially Snap packages?
On Thursday 16 Jun 2016 21:25:01 J. García wrote: > El jue, 16-06-2016 a las 19:40 -0400, José Maldonado escribió: > > That is possible, but the goal is to serve Snap container for > > applications that can be downloaded and used by the user, down a > > single > > binary that will have all the dependencies in that binary. Docker and > > LXC obviously can do this, but its scope and possibilities are much > > larger and are not addressed within the scope of normal user of a PC. > > > > > > Docker doesn't get the applications down to a single binary, it's a > package containing everything. A single binary would be something like > what Go does by default, as it compiles every source package imported > into the final binary, that's why even a "hello world" takes ~2MB. > > > > > > > > > > > > > > [AFAIK, Flatpak's for GUI apps accessed via Gnome Software so it's > > > not > > > quite a Snap competitor.] > > > > > > > > They say it's not a GNOME thing only, but born in the GNOME project, > Quote from their FAQ: > > "Is Flatpak tied to GNOME? > > No. While Flatpak has been developed by people with a long involvement > in the GNOME community it is not tied to any desktop. In fact, it was > designed with the explicit goal of allowing it to build applications > using any library stack or programming language an application author > might want." > > I would say is the implementation of something that Lennart P. wrote in > his blog a while back[1](I don't know to what extent is 'his' idea, or > if it just happens that he wrote about it after discussing it with > others), but it seems that he didn't write code for it(I looked at the > contributors in GitHub) > > > Flatpak and Snap, have GUI and command-line. In addition, Flatpak > > packages weigh less than their counterparts Snap, and right now > > several > > free software projects officially support it, including LibreOffice. > > > > > > The flatpak packages take less space because there's a separation > between runtimes and applications, with the runtime(s) containing many > of the libraries/packages required by an application, and intended to > be used by many of these, and the application package only containing > the remaining required libraries, or maybe only the app, so it could > reduce but not eliminate the problem previously discussed of > dependencies being left unmaintained and not upgraded with security > fixes. IMHO Flatpak seems a better option than Snap, and certainly > reducing file system and device access is a good thing about both, but > with these advantages some other problems are created, so it's a trade- > off. > As Andrew Savchenko said previously Snap seems like C:\Program Files > for Linux, but I would add 'with sandboxing' and other security > features, and that certainly makes it better than than Windows to be > fair. > Maybe we will see Snaps/Flatpaks of popular proprietary software that's > only available for Windows and MacOS right now that has no real FOSS > competitor e.g. AutoCAD and family, I often hear the excuse of these > vendors not supporting Linux because of the many distributions. Getting > LibreCAD to the level of AutoCAD would take a decade or more at the > pace it is going, right know it reminds me of AutoCAD 2004, and it > isn't even a that level. Trying to be optimistic maybe we'll see a new > wave of users in Linux as a result of these new packaging systems, and > in the long run if the GNU/Linux user base grows and learns about the > Free Software philosophy and get tired of having to pay large sums of > money to Autodesk and other companies for a yearly permission to use > their software, they would contribute to the FOSS alternatives with > money to get people working full time on these, and we could see them > grow to be real competitors. > That said I hope upstreams don't start bundling libraries into their > software as a result of this(at least not more than some already do > now), that's really annoying and it could create a nightmare of the > likes of java(I mean most java developers seemingly putting every jar > they come across in their 'source' trees and then forget about it for > the rest of their lifes, or at least until Oracle breaks them, after > years and years of deprecation). > > [1] http://0pointer.net/blog/revisiting-how-we-put-together-linux-syste > ms.html > How does Nix compare to flatpack, docker, snap, et al. from a gentoo perspective? https://nixos.org/nix/about.html -- Regards, Mick signature.asc Description: This is a digitally signed message part.
[gentoo-user] Virtual machines, and creating bootable images
I am looking at creating a gentoo install which I can use for my hosted servers. The documentation I have seen about doing so is to mount an "ISO" and then dd the mounted device to a data drive, make the data drive bootable (set it in the settings for hosting provider) and attach it to a server. I want to be able to build these images and test them in a virtual machine on my local server, then end up doing what I need on the hosting provider to get them in place. Quick research hasn't given me any clear guidance as to how I can export the virtual machine (ie. what even can do this) to an image that can be copied to a drive to produce a bootable drive. Anyone have some good starting points to think about this? Jigme Datse Yli-Rasku -- Jigme Datse Yli-Rasku jigme.da...@datsemultimedia.com (Preferred address for new messages) 250-505-6117 Jigme Datse Yli-Rasku PO Box 270 Rossland, BC V0G 1Y0 Canada ... ... This message should be electronically signed, and if the sender ... ... has your public key, may also be encrypted. ... ... If you have any questions about this, please email, or call.... ... ... ... Note, unknown calls likely will go to voicemail.... ... Please leave a message if you get voicemail.... ... signature.asc Description: OpenPGP digital signature