From vmoutoussamy at suse.com Mon Sep 21 08:52:46 2015 From: vmoutoussamy at suse.com (Vincent Moutoussamy) Date: Mon, 21 Sep 2015 16:52:46 +0200 Subject: [sles-beta] [sled-beta][sleha-beta]SLE12 SP1 Beta call Message-ID: <3DF03BBF-F34F-4F1F-98A3-4D21D17428D9@suse.com> Hi everyone, We would like to invite you to our SLE12 SP1 Beta Call conference scheduled for Thursday, September 24th 2015, 5:00 PM - 6:30 PM CEST. == Beta Call Topic == The main purpose of this Beta Call conference is the presentation of the Service Pack Migration from SLE12 to SLE12 SP1. Of course we will have an interactive session to retrieve your feedback about the SLE 12 SP1 Beta Program and to respond to your questions. Thorsten Kukuk our Senior Architect SLES, will be the presenter and will cast the following presentation. The Service Pack Migration from SLE12 to SLE12 SP1: - History/why we change it - What is supported - Definitions/Workflow - Own/Third Party Repositories - Migration via YaST2 - Migration via zypper - Migration via own scripts - Rollback/Recovery == When/Where/How == SUSE - SLE12 SP1 Beta call Thursday, September 24th 2015, 5:00 PM - 6:30 PM CEST Please join our meeting from your computer, tablet or smartphone at https://global.gotomeeting.com/join/117531805 You can also dial in using your phone. United States : +1 (224) 501-3217 !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! Access Code: 117-531-805 Password: susebeta !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! More phone numbers: Australia : +61 2 9087 3605 Austria : +43 (0) 7 2088 1403 Belgium : +32 (0) 28 93 7019 Canada : +1 (647) 497-9351 Denmark : +45 (0) 69 91 88 64 Finland : +358 (0) 942 41 5780 France : +33 (0) 182 880 459 Germany : +49 (0) 692 5736 7210 Ireland : +353 (0) 14 845 978 Italy : +39 0 553 98 95 67 Netherlands : +31 (0) 208 080 381 New Zealand : +64 4 974 7214 Norway : +47 21 03 58 98 Spain : +34 955 32 0845 Sweden : +46 (0) 853 527 836 Switzerland : +41 (0) 435 0167 09 United Kingdom : +44 (0) 330 221 0086 == Misc == - The presentation will be recorded so if you can?t attend to the Beta Call, you will be able to watch it later. We hope to share the video with you as soon as possible. - If you have any issues to join the Beta Call, or during the conference and if you have any questions, please contact us at beta-programs at lists.suse.com or use the built-in chat during the conference. - Firefox and Chrome should both work fine for this conference, please try one or the other if you experience any trouble. Regards, Your SUSE Linux Enterprise Team -- Vincent Moutoussamy SUSE Beta Program and SDK Project Manager From behlert at suse.com Thu Sep 24 09:15:32 2015 From: behlert at suse.com (Stefan Behlert) Date: Thu, 24 Sep 2015 17:15:32 +0200 Subject: [sles-beta] SLE 12 SP1 RC1: most likely on Friday Message-ID: <20150924151532.GN10160@suse.de> Hi all, We know you are all waiting for RC1 of SP1 to be released. We have given images to our QA meanwhile, and the tests look good until now. QA testing is not finished yet, but from the current testing progress the expectations is that RC1 will be available tomorrow. Of course, this depepends on QA not finding new bugs,, which we currently do not expect ;) thanks, Stefan, for the Release Managers -- Stefan Behlert, SUSE LINUX Release Manager Enterprise Server Maxfeldstr. 5, D-90409 Nuernberg, Germany Phone +49-911-74053-173 SUSE Linux GmbH, GF: Felix Imend?rffer, Jane Smithard, Graham Norton, HRB 21284 (AG N?rnberg) From mahendral at millenniumit.com Fri Sep 25 04:50:23 2015 From: mahendral at millenniumit.com (Laxman, Mahendra) Date: Fri, 25 Sep 2015 10:50:23 +0000 Subject: [sles-beta] SLES 12 SP1 BETA3 - tickless kernel In-Reply-To: <20150925095059.GA7973@fm.suse.cz> References: <049A43DFF32DB14DBA34752BE3659AED58D2BF9E@LG-SPMB-MBX01.lseg.stockex.local>, <20150925095059.GA7973@fm.suse.cz> Message-ID: <049A43DFF32DB14DBA34752BE3659AED58D2C0F5@LG-SPMB-MBX01.lseg.stockex.local> Hi Libor, It is great to here about that.. CONFIG_NO_HZ_FULL=y will also enhance power-saving as well. I think it will be beneficial to have it in generic kernel as well. Is there any reason that we can't have this in generic kernel? BTW RHEL generic kernel already having this enabled. Regards, Mahendra. ________________________________________ From: Libor Pechacek [lpechacek at suse.com] Sent: 25 September 2015 15:20 To: Laxman, Mahendra Cc: sles-beta at lists.suse.com Subject: Re: [sles-beta] SLES 12 SP1 BETA3 - tickless kernel Hello Mahendra, A kernel with CONFIG_NO_HZ_FULL=y enabled is part of the upcoming SLERT 12 SP1. Lab test results with sysjitter and the latest SLERT development snapshot show equally good results as your ones. Regards, Libor Pechacek On Fri 25-09-15 06:22:13, Laxman, Mahendra wrote: > Hi all, > > I would like to have tickless mode enable in SLES12 SP1 kernel to reduce the > system jitter produced by the clock interrupts. > I did test this behavior on the kernel version 3.12.47-2-default and it > reduced a significant amount of jitter. > The jitter was measured using sysjitter-1.3 > [http://www.openonload.org/download.html] with command './sysjitter --runtime > 20 300'. > Only the last line of the test result is given below which represent the > percentage jitter caused through 20s test duration and the lower values are > better. > > without tickless mode (default Kernel with boot parameters: isolcpus=1-15 ) > core_i: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 > int_total(%): 0.152 0.113 0.113 0.113 0.116 0.115 0.113 0.116 0.140 0.120 0.122 0.119 0.120 0.122 0.119 0.119 > > with tickless mode (Kernel compiled with CONFIG_NO_HZ_FULL=y and boot parameters: nohz_full=1-15 isolcpus=1-15 rcu_nocbs=1-15) > core_i: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 > int_total(%): 0.365 0.015 0.017 0.015 0.019 0.015 0.015 0.019 0.004 0.004 0.004 0.004 0.003 0.004 0.004 0.004 > > Full test output is also attached herewith. > > > What is the possibility of having tickless mode enabled? > > > Thanks & Regards, > Mahendra. -- Libor Pechacek SUSE Labs This e-mail transmission (inclusive of any attachments) is strictly confidential and intended solely for the ordinary user of the e-mail address to which it was addressed. It may contain legally privileged and/or CONFIDENTIAL information. The unauthorized use, disclosure, distribution printing and/or copying of this e-mail or any information it contains is prohibited and could, in certain circumstances, constitute an offence. If you have received this e-mail in error or are not an intended recipient please inform the sender of the email and MillenniumIT immediately by return e-mail or telephone (+94-11) 2416000. We advise that in keeping with good computing practice, the recipient of this e-mail should ensure that it is virus free. We do not accept responsibility for any virus that may be transferred by way of this e-mail. E-mail may be susceptible to data corruption, interception and unauthorized amendment, and we do not accept liability for any such corruption, interception or amendment or any consequences thereof. www.millenniumit.com From anijhawan at vmware.com Fri Sep 25 08:23:17 2015 From: anijhawan at vmware.com (Aman Nijhawan) Date: Fri, 25 Sep 2015 14:23:17 +0000 Subject: [sles-beta] systemd sysvinit-compat weirdness Message-ID: <081EC633-D8A6-4311-B723-85E8D9AE09B7@vmware.com> I am seeing lockups on shutdown and reboot in sles12 with my services on sles12 that are using sysvinit-compat. It seems systemd is not shutting down services correctly. My solution is to write a custom script that gets invoked before the shutdown target and before the reboot target and shuts down the service correctly on reboot. But this feels like a hack and also slows the shutdown process down. Is anyone else seeing this? Is there a better solution? This clearly seems like a systemd-sysvinit compat bug to me. Thanks Aman From lpechacek at suse.com Fri Sep 25 08:57:46 2015 From: lpechacek at suse.com (Libor Pechacek) Date: Fri, 25 Sep 2015 16:57:46 +0200 Subject: [sles-beta] SLES 12 SP1 BETA3 - tickless kernel In-Reply-To: <049A43DFF32DB14DBA34752BE3659AED58D2C0F5@LG-SPMB-MBX01.lseg.stockex.local> References: <049A43DFF32DB14DBA34752BE3659AED58D2BF9E@LG-SPMB-MBX01.lseg.stockex.local> <20150925095059.GA7973@fm.suse.cz> <049A43DFF32DB14DBA34752BE3659AED58D2C0F5@LG-SPMB-MBX01.lseg.stockex.local> Message-ID: <20150925145743.GA4898@fmn.suse.cz> Mahendra, The main reason for not enabling CONFIG_NO_HZ_FULL in SLES 12 SP1 by default is immaturity of the code in v3.12. There are 360 additional patches needed to get RCU, timers, workqueues and other key subsystems behave well in adaptive tick mode. Such a large backport was out of scope for the service pack. Ad power saving, adaptive tick mode triggers for CPUs which have only one task to run. Chances are the task bound to the CPU is running (otherwise we wouldn't be interested in uninterrupted execution, would we?) hence periodic timer interrupt will not increase current drawn by the CPU. Last but not least, the feature is not for free. Increased code complexity causes ~0.5% syscall slow down with nohz_full=off. This performance penalty increases to 100% when nohz_full is enabled. This is also something we will take into account when enabling this feature in stock SLES kernel. Hope this provides necessary context. Regards, Libor Pechacek On Fri 25-09-15 10:50:23, Laxman, Mahendra wrote: > Hi Libor, > > It is great to here about that.. > > CONFIG_NO_HZ_FULL=y will also enhance power-saving as well. > I think it will be beneficial to have it in generic kernel as well. > Is there any reason that we can't have this in generic kernel? > > BTW RHEL generic kernel already having this enabled. > > Regards, > Mahendra. > ________________________________________ > From: Libor Pechacek [lpechacek at suse.com] > Sent: 25 September 2015 15:20 > To: Laxman, Mahendra > Cc: sles-beta at lists.suse.com > Subject: Re: [sles-beta] SLES 12 SP1 BETA3 - tickless kernel > > Hello Mahendra, > > A kernel with CONFIG_NO_HZ_FULL=y enabled is part of the upcoming SLERT 12 SP1. > Lab test results with sysjitter and the latest SLERT development snapshot > show equally good results as your ones. > > Regards, > Libor Pechacek > > On Fri 25-09-15 06:22:13, Laxman, Mahendra wrote: > > Hi all, > > > > I would like to have tickless mode enable in SLES12 SP1 kernel to reduce the > > system jitter produced by the clock interrupts. > > I did test this behavior on the kernel version 3.12.47-2-default and it > > reduced a significant amount of jitter. > > The jitter was measured using sysjitter-1.3 > > [http://www.openonload.org/download.html] with command './sysjitter --runtime > > 20 300'. > > Only the last line of the test result is given below which represent the > > percentage jitter caused through 20s test duration and the lower values are > > better. > > > > without tickless mode (default Kernel with boot parameters: isolcpus=1-15 ) > > core_i: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 > > int_total(%): 0.152 0.113 0.113 0.113 0.116 0.115 0.113 0.116 0.140 0.120 0.122 0.119 0.120 0.122 0.119 0.119 > > > > with tickless mode (Kernel compiled with CONFIG_NO_HZ_FULL=y and boot parameters: nohz_full=1-15 isolcpus=1-15 rcu_nocbs=1-15) > > core_i: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 > > int_total(%): 0.365 0.015 0.017 0.015 0.019 0.015 0.015 0.019 0.004 0.004 0.004 0.004 0.003 0.004 0.004 0.004 > > > > Full test output is also attached herewith. > > > > > > What is the possibility of having tickless mode enabled? > > > > > > Thanks & Regards, > > Mahendra. > > > -- > Libor Pechacek > SUSE Labs > > > This e-mail transmission (inclusive of any attachments) is strictly confidential and intended solely for the ordinary user of the e-mail address to which it was addressed. It may contain legally privileged and/or CONFIDENTIAL information. The unauthorized use, disclosure, distribution printing and/or copying of this e-mail or any information it contains is prohibited and could, in certain circumstances, constitute an offence. If you have received this e-mail in error or are not an intended recipient please inform the sender of the email and MillenniumIT immediately by return e-mail or telephone (+94-11) 2416000. We advise that in keeping with good computing practice, the recipient of this e-mail should ensure that it is virus free. We do not accept responsibility for any virus that may be transferred by way of this e-mail. E-mail may be susceptible to data corruption, interception and unauthorized amendment, and we do not accept liability for any such corruption, interception > or amendment or any consequences thereof. www.millenniumit.com > > > _______________________________________________ > sles-beta mailing list > sles-beta at lists.suse.com > http://lists.suse.com/mailman/listinfo/sles-beta -- Libor Pechacek SUSE Labs From fcrozat at suse.com Fri Sep 25 09:31:18 2015 From: fcrozat at suse.com (Frederic Crozat) Date: Fri, 25 Sep 2015 17:31:18 +0200 Subject: [sles-beta] systemd sysvinit-compat weirdness In-Reply-To: <081EC633-D8A6-4311-B723-85E8D9AE09B7@vmware.com> References: <081EC633-D8A6-4311-B723-85E8D9AE09B7@vmware.com> Message-ID: <1443195078.4788.7.camel@par-r81vxc7.par.novell.com> Le vendredi 25 septembre 2015 ? 14:23 +0000, Aman Nijhawan a ?crit : > I am seeing lockups on shutdown and reboot in sles12 with my services on sles12 that are using sysvinit-compat. > It seems systemd is not shutting down services correctly. > > My solution is to write a custom script that gets invoked before the shutdown target and before the reboot target and shuts down the service correctly on reboot. > But this feels like a hack and also slows the shutdown process down. > > Is anyone else seeing this? Is there a better solution? Usually, when people reports this kind of behavior, it is because the initscript is doing some "ugly" things or is tricking systemd in the way it is trying to map the initscript to one of the "common" system service type. So, my questions would be : - is your script tries to switch to another user (calling su foobar) to start a process as a non-root user ? - is your script tries to spawn a process which might take a long time before it terminates ? The best way to debug this would be to boot systemd into debug mode (add systemd.log_level=debug to your kernel cmdline or use "systemd-analyse set-log-level debug" to change the debug output for the current session), then start your initscript, then terminate it (I'm assuming you have the same issue when stopping it and not only at shutdown). -- Frederic Crozat Enterprise Desktop Release Manager SUSE From behlert at suse.com Fri Sep 25 09:55:02 2015 From: behlert at suse.com (Stefan Behlert) Date: Fri, 25 Sep 2015 17:55:02 +0200 Subject: [sles-beta] [ANNOUNCE] SLES 12 SP1 RC1 is available Message-ID: <20150925155502.GH10160@suse.de> !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! !! SUSE CONFIDENTIAL !! SUSE CONFIDENTIAL !! SUSE CONFIDENTIAL !! !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! Dear Beta participants, we are happy to announce the first Release Candidate of SUSE Linux Enterprise Server 12 SP1. ISO images are now available for download now. Please go to http://www.novell.com/beta and select "View my beta page". Here you should see all Beta's you are part of. Note that the directory contains the images for the SDK as well as for SLES. We offer 3 DVD ISOs: DVD1 contains the binaries, the second DVD the sources and the third DVD the debuginfo packages. The final product will not contain the debuginfo packages on the media. For installation purposes you just need Media 1 for your architecture. Please verify the sha256sum of the ISO using the SHA256SUMS file, which can be found in the same directory on the download servers. Known issues: * Cannot use silver image with openstack cloud any longer [bsc#947348] [bsc#938908] In case you see issues in a virtual image with network names, remove the file /etc/udev/rules.d/70-persistent-net.rules * Registration is not working with your regular SLE key, special regcode is required * Control characters are not interpreted in Linux console [bsc#904214] * Beta3 Intel ixgbe 4.0.1-k driver upgrade causes intermittent connectivity, unusable network [bsc#944602] With this RC1 snapshot we have reached several milestones: o Milestone: All SLE12 SP1 features have been functionally tested. o Milestone: All new/updated device drivers have been tested. o Continue install and functional tests. o Continue tests of new/updated device drivers. o Start system test. For the next milestone release we are targetting these actions and milestones: o Milestone: All translations available. o Official ISV/IHV (re-)certification/validation starts. o Official partner acceptance continues. o Only showstopper and security bugfixes get integrated. o Further security updates will be handled as separate packages. Thanks in advance for all your testing Your SUSE Linux Enterprise Team -- Stefan Behlert, SUSE LINUX Release Manager Enterprise Server Maxfeldstr. 5, D-90409 Nuernberg, Germany Phone +49-911-74053-173 SUSE Linux GmbH, GF: Felix Imend?rffer, Jane Smithard, Graham Norton, HRB 21284 (AG N?rnberg) From anijhawan at vmware.com Fri Sep 25 10:08:19 2015 From: anijhawan at vmware.com (Aman Nijhawan) Date: Fri, 25 Sep 2015 16:08:19 +0000 Subject: [sles-beta] systemd sysvinit-compat weirdness In-Reply-To: <1443195078.4788.7.camel@par-r81vxc7.par.novell.com> References: <081EC633-D8A6-4311-B723-85E8D9AE09B7@vmware.com> <1443195078.4788.7.camel@par-r81vxc7.par.novell.com> Message-ID: Actually its a list of 20 services, I have tried getting a late debug shell on the system and it seems it is not able to swap off correctly. I am not doing any special mapping from my init scripts, there are a couple of scripts doing the su but they are not the ones having an issue shutting down. This issue is not there if I manually try to call shutdown. >From my understanding of systemd-sysvinit compat. It should just call the init.d script?s start and stop does it not do that? Is there a support person you can put me in touch with that I can share my system with? Thanks Aman On Sep 25, 2015, at 8:31 AM, Frederic Crozat wrote: > Le vendredi 25 septembre 2015 ? 14:23 +0000, Aman Nijhawan a ?crit : >> I am seeing lockups on shutdown and reboot in sles12 with my services on sles12 that are using sysvinit-compat. >> It seems systemd is not shutting down services correctly. >> >> My solution is to write a custom script that gets invoked before the shutdown target and before the reboot target and shuts down the service correctly on reboot. >> But this feels like a hack and also slows the shutdown process down. >> >> Is anyone else seeing this? Is there a better solution? > > Usually, when people reports this kind of behavior, it is because the > initscript is doing some "ugly" things or is tricking systemd in the way > it is trying to map the initscript to one of the "common" system service > type. > > So, my questions would be : > - is your script tries to switch to another user (calling su foobar) to > start a process as a non-root user ? > - is your script tries to spawn a process which might take a long time > before it terminates ? > > The best way to debug this would be to boot systemd into debug mode (add > systemd.log_level=debug to your kernel cmdline or use "systemd-analyse > set-log-level debug" to change the debug output for the current > session), then start your initscript, then terminate it (I'm assuming > you have the same issue when stopping it and not only at shutdown). > > > -- > Frederic Crozat > Enterprise Desktop Release Manager > SUSE > > > _______________________________________________ > sles-beta mailing list > sles-beta at lists.suse.com > https://urldefense.proofpoint.com/v2/url?u=http-3A__lists.suse.com_mailman_listinfo_sles-2Dbeta&d=BQIGaQ&c=Sqcl0Ez6M0X8aeM67LKIiDJAXVeAw-YihVMNtXt-uEs&r=NplWt5dPxRkI-IKERB_URshYbPCJx5RwT-hRor4bJJI&m=F8KoFFUJZjvI_N78_w7TXUz6kYe9-i1nUTL8xNhDyhA&s=9aN4DhEVdZbtB2xdtjMkxAc2w28ov07ebk2sY_yo7Hk&e= From fcrozat at suse.com Fri Sep 25 10:24:07 2015 From: fcrozat at suse.com (Frederic Crozat) Date: Fri, 25 Sep 2015 18:24:07 +0200 Subject: [sles-beta] systemd sysvinit-compat weirdness In-Reply-To: References: <081EC633-D8A6-4311-B723-85E8D9AE09B7@vmware.com> <1443195078.4788.7.camel@par-r81vxc7.par.novell.com> Message-ID: <1443198247.4788.13.camel@par-r81vxc7.par.novell.com> Le vendredi 25 septembre 2015 ? 16:08 +0000, Aman Nijhawan a ?crit : > Actually its a list of 20 services, I have tried getting a late debug shell on the system and it seems it is not able to swap off correctly. > > I am not doing any special mapping from my init scripts, there are a couple of scripts doing the su but they are not the ones having an issue shutting down. > > This issue is not there if I manually try to call shutdown. > > From my understanding of systemd-sysvinit compat. It should just call the init.d script?s start and stop does it not do that? Not exactly. Systemd will still monitor PIDs from each service (including the one coming from initscripts) and will ensure they are properly terminate before considering the initscript as "finished". > Is there a support person you can put me in touch with that I can share my system with? Please open a sr, so we get this handled "properly". > Thanks > Aman > On Sep 25, 2015, at 8:31 AM, Frederic Crozat wrote: > > > Le vendredi 25 septembre 2015 ? 14:23 +0000, Aman Nijhawan a ?crit : > >> I am seeing lockups on shutdown and reboot in sles12 with my services on sles12 that are using sysvinit-compat. > >> It seems systemd is not shutting down services correctly. > >> > >> My solution is to write a custom script that gets invoked before the shutdown target and before the reboot target and shuts down the service correctly on reboot. > >> But this feels like a hack and also slows the shutdown process down. > >> > >> Is anyone else seeing this? Is there a better solution? > > > > Usually, when people reports this kind of behavior, it is because the > > initscript is doing some "ugly" things or is tricking systemd in the way > > it is trying to map the initscript to one of the "common" system service > > type. > > > > So, my questions would be : > > - is your script tries to switch to another user (calling su foobar) to > > start a process as a non-root user ? > > - is your script tries to spawn a process which might take a long time > > before it terminates ? > > > > The best way to debug this would be to boot systemd into debug mode (add > > systemd.log_level=debug to your kernel cmdline or use "systemd-analyse > > set-log-level debug" to change the debug output for the current > > session), then start your initscript, then terminate it (I'm assuming > > you have the same issue when stopping it and not only at shutdown). > > > > > > -- > > Frederic Crozat > > Enterprise Desktop Release Manager > > SUSE > > > > > > _______________________________________________ > > sles-beta mailing list > > sles-beta at lists.suse.com > > https://urldefense.proofpoint.com/v2/url?u=http-3A__lists.suse.com_mailman_listinfo_sles-2Dbeta&d=BQIGaQ&c=Sqcl0Ez6M0X8aeM67LKIiDJAXVeAw-YihVMNtXt-uEs&r=NplWt5dPxRkI-IKERB_URshYbPCJx5RwT-hRor4bJJI&m=F8KoFFUJZjvI_N78_w7TXUz6kYe9-i1nUTL8xNhDyhA&s=9aN4DhEVdZbtB2xdtjMkxAc2w28ov07ebk2sY_yo7Hk&e= > > -- Frederic Crozat Enterprise Desktop Release Manager SUSE