Re: [systemd-devel] [networkd] Mixing DHCP & static IPs on 1 interface
Am 22.02.2017 um 01:30 schrieb Ian Pilcher: I'm trying to find a way to do this with systemd-networkd. The reason is that my cable modem listens on a 192.168.X.X address. Normally this "just works". My firewall tries to send traffic destined for this address to my ISP's router, and the cable modem intercepts the packets and responds. If I lose connectivity, however, my firewall doesn't have a default route, so it doesn't know where to send packets destined for 192.168.X.X. The net result is that I lose connectivity to my cable modem's diagnostic pages at exactly the time that I need to access them. (OK, I don't really lose connectivity; I just have to manually add an IP address on the proper subnet to the firewall's external interface. It works, but it's so ... MANUAL! :-) since this should be all on the LAN side something is *very* unusual on your setup - the firewall i setup at office is just a virtual machine with a single NIC and is able to do NAT and filtering as well as traffic-shaping (limit all workstations together to 80% of the WAN line) with a single IP address - so i don't see any reason why your firewall can't forward packages to your router independent of the WAN state on the other side of the router ___ systemd-devel mailing list systemd-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/systemd-devel
[systemd-devel] [networkd] Mixing DHCP & static IPs on 1 interface
I'm trying to find a way to do this with systemd-networkd. The reason is that my cable modem listens on a 192.168.X.X address. Normally this "just works". My firewall tries to send traffic destined for this address to my ISP's router, and the cable modem intercepts the packets and responds. If I lose connectivity, however, my firewall doesn't have a default route, so it doesn't know where to send packets destined for 192.168.X.X. The net result is that I lose connectivity to my cable modem's diagnostic pages at exactly the time that I need to access them. (OK, I don't really lose connectivity; I just have to manually add an IP address on the proper subnet to the firewall's external interface. It works, but it's so ... MANUAL! :-) My goal is to have both the DHCP assigned address (from my ISP) and the static address always configured on the external interface. I've tried creating two separate .network files that match the interface, but only the DHCP address is getting assigned. (The old network service actually is able to set this up on boot, but the static IP eventually goes away. I suspect that dhclient is deleting it when it renews its lease.) Any ideas? -- Ian Pilcher arequip...@gmail.com "I grew up before Mark Zuckerberg invented friendship" ___ systemd-devel mailing list systemd-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/systemd-devel
[systemd-devel] v233 around the corner
Heya! just a short heads-up: we are currently working on the last polishing of v233, and hope to release it shortly. Please give current git master a whirl in preparation. Thanks, Lennart -- Lennart Poettering, Red Hat ___ systemd-devel mailing list systemd-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/systemd-devel
Re: [systemd-devel] Does systemd compression makes use of multi core
On Thu, 12.01.17 12:09, P.R.Dinesh (pr.din...@gmail.com) wrote: > Hi, > > We are using systemd version 229. Our processor is x86 octa core. > I have configured systemd coredump (coredump.conf) for storage:external and > compression:yes. > > One of our daemon consumes lots of ram memory (Around 10Gb). When it got > crashed, systemd-coredump utility took almost three hours( I have disabled > the resource limitation) to save the compressed coredump. It is > compressing to xz format. > > How could i improve this time taken for compression. Does systemd coredump > compression utility makes use of the octa core? or does it uses only a > single core? Is there any other way to handle this case better? We are using liblzma and do not use any special settings with that. If that library can do multicore compression we use it. If not, we don't. Lennart -- Lennart Poettering, Red Hat ___ systemd-devel mailing list systemd-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/systemd-devel
Re: [systemd-devel] nspawn: premounted /dev, /proc, /sys ?
On Mon, 09.01.17 23:02, Seraphime Kirkovski (kirkser...@gmail.com) wrote: > Hello systemd devs ! > > Having nspawn run containers with premounted /dev, /proc and /sys has > been on the TODO list for a while now. I'm willing to write a patch for > this as my recent usage of nspawn requires it in some sense. > > Before starting, I would like to know if there isn't already someone > working on it in the community. > > And more importantly I would like to know if an nspawn capable of > running containers with premounted and prepopulated /dev, /proc or /sys > would be an acceptable thing to have in upstream systemd. I'm asking > this because I saw recently 8274e40d8: > > nspawn: reword notice when /dev is pre-mounted and populated > (#4971) > > where a new error message suggests it is possible to use a premounted, > but not prepopulated /dev. I don't quite understand why using > a premounted, but unpopulated /dev would be useful, but the message let > me think that the opposite is probably undesired. I think pre-mounting /dev, /proc and /sys should actually work fine (at least on git), as long as you set them up properly, and in the case of /dev populate them. YMMV though if you enable userns, as then /proc and /sys should be mounted from within the userns to get the right perms. Lennart -- Lennart Poettering, Red Hat ___ systemd-devel mailing list systemd-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/systemd-devel
Re: [systemd-devel] Renew DHCP lease
On Fri, 10.02.17 16:55, John Lane (syst...@jelmail.com) wrote: > I've just switched a server to systemd-networkd dhcp client. > > How do I renew a lease and/or force it to push a new hostname to the > dhcp server. > > I have RTFM and get the impression it can't be done. Hopefully I have > that wrong... This is currently not implemented. Please file an RFE bug on github. Lennart -- Lennart Poettering, Red Hat ___ systemd-devel mailing list systemd-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/systemd-devel
Re: [systemd-devel] [RFC] systemd-swap
On Mon, 20.02.17 20:02, Timofey Titovets (nefelim...@gmail.com) wrote: > Hi, some time ago (about 3 years ago, first commit at Apr 9, 2014) > I created small project: systemd-swap. > https://github.com/Nefelim4ag/systemd-swap > > Short description from README.md: > Script for manage swap on: > zswap - Enable/Configure > zram - Autoconfigurating > files - (sparse files for saving space, support btrfs) > block devices - auto find and do swapon > It is configurable in /etc/systemd/swap.conf. > > (Project writen on pure bash) > > The main reason why i create it: > I want have a one tool to manage all possible swap spaces, with small > as possible user interaction. > > I can't collect full statistic about usage, as not so many people send > a bad feedbacks, > i only can say it's packaged for Archlinux: > https://www.archlinux.org/packages/community/any/systemd-swap/ > And that i have some "stars" and forks on github. > > For what it can be used: > I know that many users use it for create Zram + swap on their notebooks > I know that someone use it to create some types of swap files. > I use it on my PC for automount swap partitions and on my production > VMs for manage swap spaces. > > Now, for me, project looks like "enough mature to move on", and i > think about attempt to rewrite script on C, and merge it to main > systemd repository. > > But before this, i like to get some feedback from systemd developers: > Did you think it's useful enough to be rewriten and merged in > systemd repo? So, we are conservative with adding completely new components to the main systemd repo these days, and prefer instead to add new stuff into seperate projects under the "systemd" umbrella group on github. This particularly applies to stuff that isn't necessarily at the core where computing will go to, and quire frankly I have the suspicion that swap partitions and suchlike are on their way out, and are not going to be the default how future systems are set up. Hence, yeah, we are happy to open up the github systemd umbrella group, and add systemd-swap to it (well, if it follows our general coding guidelines at least to some level, i.e. is written and C and so on), but I am not sure this really belongs in the main systemd repo itself. I hope this makes sense? Lennart -- Lennart Poettering, Red Hat ___ systemd-devel mailing list systemd-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/systemd-devel
[systemd-devel] SCHED_FIFO
hi I have ctdb use team net devices but there is a problem: https://bugzilla.redhat.com/show_bug.cgi?id=1421378 2017/02/21 15:20:34.923629 [16515]: Unable to set scheduler to SCHED_FIFO (Operation not permitted) I do: system edit ctdb there: [Service] ExecStartPre=/bin/sh -c "echo 1 > /sys/fs/cgroup/cpu/system.slice/cpu.rt_runtime_us" #LimitRTPRIO=infinity #LimitRTTIME=10 TimeoutStartSec=300 Restart=on-failure RestartSec=20 But Limit options don't work for me. Would you know if it can be tweaked with systemd's help? thx, L ___ systemd-devel mailing list systemd-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/systemd-devel
Re: [systemd-devel] Log output is incomplete or unavailable
On Tue, 21.02.17 10:00, Patrick B (patrickbake...@gmail.com) wrote: > Hi guys, > > I'm running Ubuntu 16.04.01 Server and just realised some log files are not > being created. > > Such as: > > > /var/log/kern.log > /var/log/postgresql/postgresql.log These files are created by syslog or postgres itself, are you sure you have that installed? Note that journald won't write these files. Please contact your downstream distro for help. Lennart -- Lennart Poettering, Red Hat ___ systemd-devel mailing list systemd-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/systemd-devel
Re: [systemd-devel] Systemctl causes Spark native thread creation issue
On Mon, 20.02.17 16:44, Rao Vz (raoa...@gmail.com) wrote: > Hi, Guys > > We have a Apache Spark cluster of 3 nodes, one is master and slave, the > other two are slaves. When starting Spark worker with "systemctl start > spark-worker", when running out apps, sometimes but not always it generates > "java.lang.OutOfMemoryError: unable to create new native thread" error in > Spark worker logs. I figure the error is misleading and is not about memory at all, and you need to bump the default TasksMax= field or even turn it off by setting it to infinity. Lennart -- Lennart Poettering, Red Hat ___ systemd-devel mailing list systemd-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/systemd-devel
