Re: [systemd-devel] centos7/systemd/mulitpathd
Will do sir apologies was looking if people had experienced anything like this with systemd and/or proper places to insert a pause/slow down in a unit file to let devices get populated. These volumes mounted reliable under sysint and centos 6u5. Basically was looking to see if I have to update my thinking/or gotchas for leveraging under a systemd based boot up thanks very much On Thu, Mar 30, 2017 at 5:37 AM, Lennart Poettering <lenn...@poettering.net> wrote: > On Mon, 13.03.17 16:25, jsl6uy js16uy (js1...@gmail.com) wrote: > > > Hello all, hope all is well > > Is there a dependency or delay that needs to be added to have volumes > > attached via HBA and managed by multipathd/devmapper and lvm reliably > mount > > on boot? > > Please contact the multipath/DM folks about this, the systemd > community is not the best place to look for storage advice. > > Sorry, > > 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] centos7/systemd/mulitpathd
Apologies, pinging again, to see if anyone has had issues with systemd+mutlipathd running on CentOS7u2 and the creation of devices and setting up /dev/disk/by-uuid id links during boot up. if not, no worries and have great days, will check in elsewhere thanks and regards On Mon, Mar 13, 2017 at 4:25 PM, jsl6uy js16uy <js1...@gmail.com> wrote: > Hello all, hope all is well > Is there a dependency or delay that needs to be added to have volumes > attached via HBA and managed by multipathd/devmapper and lvm reliably mount > on boot? > > More detail > > We have some volumes connected via an hba in a Dell R630 running centos 7u2 > When booting up, the volumes do not mount. Causing the server to go into > emergency mode, upon login we could see that under /dev/disk/by-uuid/ sym > links were not created. At this point we could run from the emergency > maintenance shell, partprobe or restart systemd-udev-trigger.service. This > would fully create devices under /dev/ and fully populate the links in > /dev/disk/by-uuid. Then we run mount -a. and exit the emergency shell and > server completes boot > multipath is installed and enabled > > Server is in prod so I can't do further reboot testing till the weekend. > How we have it boot now is treating the volumes as x-systemd.automount with > noauto/nofail options. After an initial "ls" and/or "cd" the volumes are > available and this prevents halts during reboots > > > workaround: > UUID=094f2d03-87a1-4c0f-9777-41d11d7e6a7d /data03 xfs > noauto,nofail,x-systemd.automount,nobarrier,swalloc, > inode64,noatime,nodiratime,logbufs=8,allocsize=16m 0 2 > Do I need a _netdev? Apologies not sure with systemd at the helm > > > thanks for any/all help > > ___ systemd-devel mailing list systemd-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/systemd-devel
[systemd-devel] centos7/systemd/mulitpathd
Hello all, hope all is well Is there a dependency or delay that needs to be added to have volumes attached via HBA and managed by multipathd/devmapper and lvm reliably mount on boot? More detail We have some volumes connected via an hba in a Dell R630 running centos 7u2 When booting up, the volumes do not mount. Causing the server to go into emergency mode, upon login we could see that under /dev/disk/by-uuid/ sym links were not created. At this point we could run from the emergency maintenance shell, partprobe or restart systemd-udev-trigger.service. This would fully create devices under /dev/ and fully populate the links in /dev/disk/by-uuid. Then we run mount -a. and exit the emergency shell and server completes boot multipath is installed and enabled Server is in prod so I can't do further reboot testing till the weekend. How we have it boot now is treating the volumes as x-systemd.automount with noauto/nofail options. After an initial "ls" and/or "cd" the volumes are available and this prevents halts during reboots workaround: UUID=094f2d03-87a1-4c0f-9777-41d11d7e6a7d /data03 xfs noauto,nofail,x-systemd.automount,nobarrier,swalloc,inode64,noatime,nodiratime,logbufs=8,allocsize=16m 0 2 Do I need a _netdev? Apologies not sure with systemd at the helm thanks for any/all help ___ systemd-devel mailing list systemd-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/systemd-devel
Re: [systemd-devel] controlling IO from a transient service
That definitely did the trick I am creating/testing local qemu-based vagrant images. Very Nice not to have my host and ssd crushed every time I add boxes to vagrant! thanks very much again sir much appreciated On Wed, Jun 29, 2016 at 4:55 PM, jsl6uy js16uy <js1...@gmail.com> wrote: > Did NOT even consider the consequence of the su. been pounding the head. > Thanks very much sir! > will try as you recommend and report back > > best regards > > On Wed, Jun 29, 2016 at 4:00 PM, Lennart Poettering < > lenn...@poettering.net> wrote: > >> On Tue, 28.06.16 11:20, jsl6uy js16uy (js1...@gmail.com) wrote: >> >> > Hello all, hope all is well >> > >> > Should I be able to control the IO for a tar process? >> > >> > I am on arch linux using systemd-run to initiate a service to which I >> can >> > apply resource limits >> > >> > Linux 4.6.2-1-ARCH #1 SMP PREEMPT Wed Jun 8 08:40:59 CEST 2016 x86_64 >> > GNU/Linux >> > >> > systemctl --version >> > systemd 230 >> > +PAM -AUDIT -SELINUX -IMA -APPARMOR +SMACK -SYSVINIT +UTMP >> +LIBCRYPTSETUP >> > +GCRYPT +GNUTLS +ACL +XZ +LZ4 +SECCOMP +BLKID +ELFUTILS +KMOD +IDN >> > >> > >> > I am starting a transient service with the parameters below. I have >> tried >> > several combinations >> > >> > sudo systemd-run --unit=loadio -p "BlockIOWriteBandwidth=/dev/dm-1 50M" >> -- >> > su - foo -c "cd $PWD && tar vcf custom_box.box ./metadata.json >> > ./Vagrantfile ./box.img" >> >> So this creates a transient service unit, and then runs "su" inside of >> it, which will create a separate transient scope unit for the login >> session "su" creates. >> >> Or in other words: you actually create a the unit correctly, but then >> immediately escape it again for the actual command you want to run. >> >> Consider using the "User=" property instead of "su". "User=" does not >> involve PAM and hence does not result in creation of a new user login >> session. >> >> 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] controlling IO from a transient service
Did NOT even consider the consequence of the su. been pounding the head. Thanks very much sir! will try as you recommend and report back best regards On Wed, Jun 29, 2016 at 4:00 PM, Lennart Poettering <lenn...@poettering.net> wrote: > On Tue, 28.06.16 11:20, jsl6uy js16uy (js1...@gmail.com) wrote: > > > Hello all, hope all is well > > > > Should I be able to control the IO for a tar process? > > > > I am on arch linux using systemd-run to initiate a service to which I can > > apply resource limits > > > > Linux 4.6.2-1-ARCH #1 SMP PREEMPT Wed Jun 8 08:40:59 CEST 2016 x86_64 > > GNU/Linux > > > > systemctl --version > > systemd 230 > > +PAM -AUDIT -SELINUX -IMA -APPARMOR +SMACK -SYSVINIT +UTMP +LIBCRYPTSETUP > > +GCRYPT +GNUTLS +ACL +XZ +LZ4 +SECCOMP +BLKID +ELFUTILS +KMOD +IDN > > > > > > I am starting a transient service with the parameters below. I have tried > > several combinations > > > > sudo systemd-run --unit=loadio -p "BlockIOWriteBandwidth=/dev/dm-1 50M" > -- > > su - foo -c "cd $PWD && tar vcf custom_box.box ./metadata.json > > ./Vagrantfile ./box.img" > > So this creates a transient service unit, and then runs "su" inside of > it, which will create a separate transient scope unit for the login > session "su" creates. > > Or in other words: you actually create a the unit correctly, but then > immediately escape it again for the actual command you want to run. > > Consider using the "User=" property instead of "su". "User=" does not > involve PAM and hence does not result in creation of a new user login > session. > > Lennart > > -- > Lennart Poettering, Red Hat > ___ systemd-devel mailing list systemd-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/systemd-devel
[systemd-devel] controlling IO from a transient service
Hello all, hope all is well Should I be able to control the IO for a tar process? I am on arch linux using systemd-run to initiate a service to which I can apply resource limits Linux 4.6.2-1-ARCH #1 SMP PREEMPT Wed Jun 8 08:40:59 CEST 2016 x86_64 GNU/Linux systemctl --version systemd 230 +PAM -AUDIT -SELINUX -IMA -APPARMOR +SMACK -SYSVINIT +UTMP +LIBCRYPTSETUP +GCRYPT +GNUTLS +ACL +XZ +LZ4 +SECCOMP +BLKID +ELFUTILS +KMOD +IDN I am starting a transient service with the parameters below. I have tried several combinations sudo systemd-run --unit=loadio -p "BlockIOWriteBandwidth=/dev/dm-1 50M" -- su - foo -c "cd $PWD && tar vcf custom_box.box ./metadata.json ./Vagrantfile ./box.img" sudo systemd-run --unit=loadio -p "IOWriteBandwidthMax=/dev/dm-1 50M" -- su - foo -c "cd $PWD && tar vcf custom_box.box ./metadata.json ./Vagrantfile ./box.img" sudo systemd-run --unit=loadio -p "IOWriteBandwidthMax=/dev/dm-1 50M" -p "BlockIOWriteBandwidth=/dev/dm-1 50M" -- su - foo -c "cd $PWD && tar vcf custom_box.box ./metadata.json ./Vagrantfile ./box.img" >>Tried both unified and legacy IO styles here After I kick off the process, I can see with iotop that writes are not being capped, and my machine becoming a bit sluggish ;). Thoughts? TIA for any/all help ___ systemd-devel mailing list systemd-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/systemd-devel
[systemd-devel] Am I getting/understanding certain bits on systemd/cgroups -- low-prio query
I dig transient services/scopes I also like interacting with resource limits via systemctl set-property interface --easy to list resources that are available and set sans finding and catting on /sys/fs/cgroup/foo So... Basically I have a system up and running. See some processes I want to corral So I would like to create a transient bucket, a scope seems best over a service with the mentality that these process were not started by a systemd unit file. so right now I run systemd-run --unit=foo --remain-after-exit cat /dev/null The transient service stays up I set an arbitrary resources limit to create files under /sys/fs/cgroup Now I move processes into the tasks file within this cgroup path created via systemctl/systemd And I now can setup limits on a group, whatever I move into the tasks file, on a set arbitrary run processes. I could just set the limits on the individual processes but would have to hit each and work with /proc and /sys/fs/cgroup to get them Corralled I really would prefer to use the structure systemd creates/leverages and use systemctl to interact with processes as a manually assembled group So does this seem correct or should I be thinking of this differently? --Also I question if I'm thinking of this right b/c of the "hack" to get an "empty" bucket that I can dump running processes into for quick resource control Apologies if this is not the right forum for a concepts question, I can try to find another list if needs be As always thanks for any/all help ___ systemd-devel mailing list systemd-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/systemd-devel
Re: [systemd-devel] issues setting limits on the fly for services/scopes in user slice
Understood thanks much sir and now at least I know I'm not crazy for that reason ;) On Fri, May 27, 2016 at 9:27 AM, Lennart Poettering <lenn...@poettering.net> wrote: > On Fri, 27.05.16 09:03, jsl6uy js16uy (js1...@gmail.com) wrote: > > > systemctl --user set-property foo.scope CPUQuota='5%' > > or > > systemctl set-property --user --runtime foo.scope CPUQuota='5%' > > This is not supported with the classic cgroups hierarchy, as resource > controller delegation to unprivileged clients is not safe. Thus, > setting these properties for "--user" instances has no effect. > > For the new cgroups-v2 hierarchy safe delegation to unprivileged > clients is finally supported, but the "cpu" cgroup resource controller > has not been ported over yet to cgroups-v2, hence this isn't available > still either yet, but will be in one of the next kernel and systemd > versions. > > Sorry, > > Lennart > > -- > Lennart Poettering, Red Hat > ___ systemd-devel mailing list systemd-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/systemd-devel
[systemd-devel] issues setting limits on the fly for services/scopes in user slice
Hello All, hope all is well On arch linux Linux X 4.5.3-1-ARCH #1 SMP PREEMPT Sat May 7 20:43:57 CEST 2016 x86_64 GNU/Linux local/libsystemd 229-3 systemd client libraries local/systemd 229-3 system and service manager local/systemd-sysvcompat 229-3 (base) sysvinit compat for systemd using.. systemd-run --user --unit=foo --scope stress -c 1 I can see the scope created under my user slice In top I can see the cpu pegged I run systemctl --user set-property foo.scope CPUQuota='5%' or systemctl set-property --user --runtime foo.scope CPUQuota='5%' or couple other postitional variations Also tried CPUShares as the controlling resource knob However the process remains at 100% usage on top Same multicore host When I set resource properties on the fly on a vm in a scope running in the machine slice launched by libvirt, I do see cpu usage drop accordingly. -Fired up an ubu vm ran stress and/or stress-ng -c1 inside the vm -systemctl set-property 'machine-qemu\x2d68\x2dchhq\x2dvulr6isi02.scope' CPUQuota='5%' -cpu usage in host drops to the expected amount CentOS Linux release 7.2.1511 (Core) On Centos 7u2 , I can't do exactly the same b/c --no user option in systemd-run --I am at the version where --slice is not working with setting a scope So I run as a service systemd-run --unit=foo --uid=2020 --gid=2020 --slice=user-2020.slice stress -c 1 systemctl set-property foo CPUQuota='5%' And I see stress hauled down to 5% usage Am I doing something wrong for on the fly process corralling that is under the user slice? I can provide more info if needed thanks for any/all help ___ systemd-devel mailing list systemd-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/systemd-devel