Re: [ceph-users] rbd-fuse Transport endpoint is not connected
It is great having access to features that are not fully production ready, but it would be nice to know which Ceph features are ready and which are not. Just like Ceph File System is well marked that it is not yet fully ready for production, it would be nice if rbd-fuse could be marked as not ready, and why. This may motivate someone in the community to work on it. Eric On Thu, Jul 30, 2015 at 2:30 AM, Ilya Dryomov idryo...@gmail.com wrote: This is an error from rados_connect(), so I'd say there is something wrong with your ceph.conf or the environment. We can try to debug this, but yeah, rbd-fuse, as it is, is not intended for serious use. Thanks, Ilya ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
[ceph-users] dropping old distros: el6, precise 12.04, debian wheezy?
As time marches on it becomes increasingly difficult to maintain proper builds and packages for older distros. For example, as we make the systemd transition, maintaining the kludgey sysvinit and udev support for centos6/rhel6 is a pain in the butt and eats up time and energy to maintain and test that we could be spending doing more useful work. Dropping them would mean: - Ongoing development on master (and future versions like infernalis and jewel) would not be tested on these distros. - We would stop building upstream release packages on ceph.com for new releases. - We would probably continue building hammer and firefly packages for future bugfix point releases. - The downstream distros would probably continue to package them, but the burden would be on them. For example, if Ubuntu wanted to ship Jewel on precise 12.04, they could, but they'd probably need to futz with the packaging and/or build environment to make it work. So... given that, I'd like to gauge user interest in these old distros. Specifically, CentOS6 / RHEL6 Ubuntu precise 12.04 Debian wheezy Would anyone miss them? In particular, dropping these three would mean we could drop sysvinit entirely and focus on systemd (and continue maintaining the existing upstart files for just a bit longer). That would be a relief. (The sysvinit files wouldn't go away in the source tree, but we wouldn't worry about packaging and testing them properly.) Thanks! sage ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] dropping old distros: el6, precise 12.04, debian wheezy?
Not at all. We have this: http://ceph.com/docs/master/releases/ I would expect that whatever distribution I install Ceph LTS release on will be supported for the time specified. That means if I install Hammer on CentOS 6 now it will stay supported until 3Q/2016. Of course if in the meantime the distribution itself becomes unsupported then it makes sense to stop supporting it for Ceph as well, but that’s probably not the case here: https://access.redhat.com/support/policy/updates/errata I don’t expect Ceph to be supported until EOL of the distro. Jan On 30 Jul 2015, at 16:34, Handzik, Joe joseph.t.hand...@hp.com wrote: So, essentially, you'd vote that all LTS/enterprise releases be supported until their vendor's (canonical, Suse, red hat) designated EOL date? Not voting either way, just trying to put a date stamp on some of this. Joe On Jul 30, 2015, at 9:30 AM, Jan “Zviratko” Schermer zvira...@zviratko.net wrote: I understand your reasons, but dropping support for LTS release like this is not right. You should lege artis support every distribution the LTS release could have ever been installed on - that’s what the LTS label is for and what we rely on once we build a project on top of it CentOS 6 in particular is still very widely used and even installed, enterprise apps rely on it to this day. Someone out there is surely maintaining their LTS Ceph release on this distro and not having tested packages will hurt badly. We don’t want out project managers selecting EMC SAN over CEPH SDS because of such uncertainty, and you should benchmark yourself to those vendors, maybe... Every developer loves dropping support and concentrating on the bleeding edge interesting stuff but that’s not how it should work. Just my 2 cents... Jan On 30 Jul 2015, at 15:54, Sage Weil sw...@redhat.com wrote: As time marches on it becomes increasingly difficult to maintain proper builds and packages for older distros. For example, as we make the systemd transition, maintaining the kludgey sysvinit and udev support for centos6/rhel6 is a pain in the butt and eats up time and energy to maintain and test that we could be spending doing more useful work. Dropping them would mean: - Ongoing development on master (and future versions like infernalis and jewel) would not be tested on these distros. - We would stop building upstream release packages on ceph.com for new releases. - We would probably continue building hammer and firefly packages for future bugfix point releases. - The downstream distros would probably continue to package them, but the burden would be on them. For example, if Ubuntu wanted to ship Jewel on precise 12.04, they could, but they'd probably need to futz with the packaging and/or build environment to make it work. So... given that, I'd like to gauge user interest in these old distros. Specifically, CentOS6 / RHEL6 Ubuntu precise 12.04 Debian wheezy Would anyone miss them? In particular, dropping these three would mean we could drop sysvinit entirely and focus on systemd (and continue maintaining the existing upstart files for just a bit longer). That would be a relief. (The sysvinit files wouldn't go away in the source tree, but we wouldn't worry about packaging and testing them properly.) Thanks! sage ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com -- To unsubscribe from this list: send the line unsubscribe ceph-devel in the body of a message to majord...@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] dropping old distros: el6, precise 12.04, debian wheezy?
Hi, much like debian already has, I would suggest to not make systemd a dependency for Ceph (or anything for that matter). The reason being here that we desperately need sysvinit until the systemd forks are ready which offer the systemd init system without all those slapped-on appendages that systemd has accumulated. Alternatively, would it possible to maintain a dedicated changelog where all changes to the init scripts are being collected so that people can maintain their own sysvinit scripts? I just can't foresee systemd becoming usable in the near future and thus we will have to rely on sysvinit for quite a while longer. RHEL6 will be maintained until Nov 2020, debian wheezy LTS until May 2018, so I feel like it would be premature to drop those already. I can't give any names but I know of at least 2 customers I am consulting for that have blocked the RHEL7 upgrade due to systemd. Regards, Marc On 07/30/2015 03:54 PM, Sage Weil wrote: As time marches on it becomes increasingly difficult to maintain proper builds and packages for older distros. For example, as we make the systemd transition, maintaining the kludgey sysvinit and udev support for centos6/rhel6 is a pain in the butt and eats up time and energy to maintain and test that we could be spending doing more useful work. Dropping them would mean: - Ongoing development on master (and future versions like infernalis and jewel) would not be tested on these distros. - We would stop building upstream release packages on ceph.com for new releases. - We would probably continue building hammer and firefly packages for future bugfix point releases. - The downstream distros would probably continue to package them, but the burden would be on them. For example, if Ubuntu wanted to ship Jewel on precise 12.04, they could, but they'd probably need to futz with the packaging and/or build environment to make it work. So... given that, I'd like to gauge user interest in these old distros. Specifically, CentOS6 / RHEL6 Ubuntu precise 12.04 Debian wheezy Would anyone miss them? In particular, dropping these three would mean we could drop sysvinit entirely and focus on systemd (and continue maintaining the existing upstart files for just a bit longer). That would be a relief. (The sysvinit files wouldn't go away in the source tree, but we wouldn't worry about packaging and testing them properly.) Thanks! sage ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] dropping old distros: el6, precise 12.04, debian wheezy?
i would certainly like that all client libs and/or kernel modules stay tested and supported on these OSes for future ceph releases. not sure how much work that is, but the at least client side shouldn't be affected by the init move. stijn On 07/30/2015 04:43 PM, Marc wrote: Hi, much like debian already has, I would suggest to not make systemd a dependency for Ceph (or anything for that matter). The reason being here that we desperately need sysvinit until the systemd forks are ready which offer the systemd init system without all those slapped-on appendages that systemd has accumulated. Alternatively, would it possible to maintain a dedicated changelog where all changes to the init scripts are being collected so that people can maintain their own sysvinit scripts? I just can't foresee systemd becoming usable in the near future and thus we will have to rely on sysvinit for quite a while longer. RHEL6 will be maintained until Nov 2020, debian wheezy LTS until May 2018, so I feel like it would be premature to drop those already. I can't give any names but I know of at least 2 customers I am consulting for that have blocked the RHEL7 upgrade due to systemd. Regards, Marc On 07/30/2015 03:54 PM, Sage Weil wrote: As time marches on it becomes increasingly difficult to maintain proper builds and packages for older distros. For example, as we make the systemd transition, maintaining the kludgey sysvinit and udev support for centos6/rhel6 is a pain in the butt and eats up time and energy to maintain and test that we could be spending doing more useful work. Dropping them would mean: - Ongoing development on master (and future versions like infernalis and jewel) would not be tested on these distros. - We would stop building upstream release packages on ceph.com for new releases. - We would probably continue building hammer and firefly packages for future bugfix point releases. - The downstream distros would probably continue to package them, but the burden would be on them. For example, if Ubuntu wanted to ship Jewel on precise 12.04, they could, but they'd probably need to futz with the packaging and/or build environment to make it work. So... given that, I'd like to gauge user interest in these old distros. Specifically, CentOS6 / RHEL6 Ubuntu precise 12.04 Debian wheezy Would anyone miss them? In particular, dropping these three would mean we could drop sysvinit entirely and focus on systemd (and continue maintaining the existing upstart files for just a bit longer). That would be a relief. (The sysvinit files wouldn't go away in the source tree, but we wouldn't worry about packaging and testing them properly.) Thanks! sage ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] dropping old distros: el6, precise 12.04, debian wheezy?
I understand your reasons, but dropping support for LTS release like this is not right. You should lege artis support every distribution the LTS release could have ever been installed on - that’s what the LTS label is for and what we rely on once we build a project on top of it CentOS 6 in particular is still very widely used and even installed, enterprise apps rely on it to this day. Someone out there is surely maintaining their LTS Ceph release on this distro and not having tested packages will hurt badly. We don’t want out project managers selecting EMC SAN over CEPH SDS because of such uncertainty, and you should benchmark yourself to those vendors, maybe... Every developer loves dropping support and concentrating on the bleeding edge interesting stuff but that’s not how it should work. Just my 2 cents... Jan On 30 Jul 2015, at 15:54, Sage Weil sw...@redhat.com wrote: As time marches on it becomes increasingly difficult to maintain proper builds and packages for older distros. For example, as we make the systemd transition, maintaining the kludgey sysvinit and udev support for centos6/rhel6 is a pain in the butt and eats up time and energy to maintain and test that we could be spending doing more useful work. Dropping them would mean: - Ongoing development on master (and future versions like infernalis and jewel) would not be tested on these distros. - We would stop building upstream release packages on ceph.com for new releases. - We would probably continue building hammer and firefly packages for future bugfix point releases. - The downstream distros would probably continue to package them, but the burden would be on them. For example, if Ubuntu wanted to ship Jewel on precise 12.04, they could, but they'd probably need to futz with the packaging and/or build environment to make it work. So... given that, I'd like to gauge user interest in these old distros. Specifically, CentOS6 / RHEL6 Ubuntu precise 12.04 Debian wheezy Would anyone miss them? In particular, dropping these three would mean we could drop sysvinit entirely and focus on systemd (and continue maintaining the existing upstart files for just a bit longer). That would be a relief. (The sysvinit files wouldn't go away in the source tree, but we wouldn't worry about packaging and testing them properly.) Thanks! sage ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] dropping old distros: el6, precise 12.04, debian wheezy?
I understand your reasons, but dropping support for LTS release like this is not right. You should lege artis support every distribution the LTS release could have ever been installed on - that’s what the LTS label is for and what we rely on once we build a project on top of it CentOS 6 in particular is still very widely used and even installed, enterprise apps rely on it to this day. Someone out there is surely maintaining their LTS Ceph release on this distro and not having tested packages will hurt badly. We don’t want out project managers selecting EMC SAN over CEPH SDS because of such uncertainty, and you should benchmark yourself to those vendors, maybe... Every developer loves dropping support and concentrating on the bleeding edge interesting stuff but that’s not how it should work. Just my 2 cents... Jan P.S. sorry if this reached your more than once, my mail client is having a very baaad day On 30 Jul 2015, at 15:54, Sage Weil sw...@redhat.com wrote: As time marches on it becomes increasingly difficult to maintain proper builds and packages for older distros. For example, as we make the systemd transition, maintaining the kludgey sysvinit and udev support for centos6/rhel6 is a pain in the butt and eats up time and energy to maintain and test that we could be spending doing more useful work. Dropping them would mean: - Ongoing development on master (and future versions like infernalis and jewel) would not be tested on these distros. - We would stop building upstream release packages on ceph.com for new releases. - We would probably continue building hammer and firefly packages for future bugfix point releases. - The downstream distros would probably continue to package them, but the burden would be on them. For example, if Ubuntu wanted to ship Jewel on precise 12.04, they could, but they'd probably need to futz with the packaging and/or build environment to make it work. So... given that, I'd like to gauge user interest in these old distros. Specifically, CentOS6 / RHEL6 Ubuntu precise 12.04 Debian wheezy Would anyone miss them? In particular, dropping these three would mean we could drop sysvinit entirely and focus on systemd (and continue maintaining the existing upstart files for just a bit longer). That would be a relief. (The sysvinit files wouldn't go away in the source tree, but we wouldn't worry about packaging and testing them properly.) Thanks! sage ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] dropping old distros: el6, precise 12.04, debian wheezy?
I'm don't prefer it. you have to maintain those releases up to their EOL. On Thu, Jul 30, 2015 at 8:48 PM, Stijn De Weirdt stijn.dewei...@ugent.be wrote: i would certainly like that all client libs and/or kernel modules stay tested and supported on these OSes for future ceph releases. not sure how much work that is, but the at least client side shouldn't be affected by the init move. stijn On 07/30/2015 04:43 PM, Marc wrote: Hi, much like debian already has, I would suggest to not make systemd a dependency for Ceph (or anything for that matter). The reason being here that we desperately need sysvinit until the systemd forks are ready which offer the systemd init system without all those slapped-on appendages that systemd has accumulated. Alternatively, would it possible to maintain a dedicated changelog where all changes to the init scripts are being collected so that people can maintain their own sysvinit scripts? I just can't foresee systemd becoming usable in the near future and thus we will have to rely on sysvinit for quite a while longer. RHEL6 will be maintained until Nov 2020, debian wheezy LTS until May 2018, so I feel like it would be premature to drop those already. I can't give any names but I know of at least 2 customers I am consulting for that have blocked the RHEL7 upgrade due to systemd. Regards, Marc On 07/30/2015 03:54 PM, Sage Weil wrote: As time marches on it becomes increasingly difficult to maintain proper builds and packages for older distros. For example, as we make the systemd transition, maintaining the kludgey sysvinit and udev support for centos6/rhel6 is a pain in the butt and eats up time and energy to maintain and test that we could be spending doing more useful work. Dropping them would mean: - Ongoing development on master (and future versions like infernalis and jewel) would not be tested on these distros. - We would stop building upstream release packages on ceph.com for new releases. - We would probably continue building hammer and firefly packages for future bugfix point releases. - The downstream distros would probably continue to package them, but the burden would be on them. For example, if Ubuntu wanted to ship Jewel on precise 12.04, they could, but they'd probably need to futz with the packaging and/or build environment to make it work. So... given that, I'd like to gauge user interest in these old distros. Specifically, CentOS6 / RHEL6 Ubuntu precise 12.04 Debian wheezy Would anyone miss them? In particular, dropping these three would mean we could drop sysvinit entirely and focus on systemd (and continue maintaining the existing upstart files for just a bit longer). That would be a relief. (The sysvinit files wouldn't go away in the source tree, but we wouldn't worry about packaging and testing them properly.) Thanks! sage ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com -- Asif Murad Khan Cell: +880-1713-114230 ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] dropping old distros: el6, precise 12.04, debian wheezy?
If hammer and firefly bugfix releases will still be packaged for these distros, I don't see a problem with this. Anyone who is operating an existing LTS deployment on CentOS 6, etc. will continue to receive fixes for said LTS release. Jon From: ceph-users on behalf of Jan “Zviratko” Schermer Date: Thursday, July 30, 2015 at 8:29 AM To: Sage Weil Cc: ceph-devel, ceph-us...@ceph.commailto:ceph-us...@ceph.com Subject: Re: [ceph-users] dropping old distros: el6, precise 12.04, debian wheezy? I understand your reasons, but dropping support for LTS release like this is not right. You should lege artis support every distribution the LTS release could have ever been installed on - that’s what the LTS label is for and what we rely on once we build a project on top of it CentOS 6 in particular is still very widely used and even installed, enterprise apps rely on it to this day. Someone out there is surely maintaining their LTS Ceph release on this distro and not having tested packages will hurt badly. We don’t want out project managers selecting EMC SAN over CEPH SDS because of such uncertainty, and you should benchmark yourself to those vendors, maybe... Every developer loves dropping support and concentrating on the bleeding edge interesting stuff but that’s not how it should work. Just my 2 cents... Jan On 30 Jul 2015, at 15:54, Sage Weil sw...@redhat.commailto:sw...@redhat.com wrote: As time marches on it becomes increasingly difficult to maintain proper builds and packages for older distros. For example, as we make the systemd transition, maintaining the kludgey sysvinit and udev support for centos6/rhel6 is a pain in the butt and eats up time and energy to maintain and test that we could be spending doing more useful work. Dropping them would mean: - Ongoing development on master (and future versions like infernalis and jewel) would not be tested on these distros. - We would stop building upstream release packages on ceph.comhttp://ceph.com for new releases. - We would probably continue building hammer and firefly packages for future bugfix point releases. - The downstream distros would probably continue to package them, but the burden would be on them. For example, if Ubuntu wanted to ship Jewel on precise 12.04, they could, but they'd probably need to futz with the packaging and/or build environment to make it work. So... given that, I'd like to gauge user interest in these old distros. Specifically, CentOS6 / RHEL6 Ubuntu precise 12.04 Debian wheezy Would anyone miss them? In particular, dropping these three would mean we could drop sysvinit entirely and focus on systemd (and continue maintaining the existing upstart files for just a bit longer). That would be a relief. (The sysvinit files wouldn't go away in the source tree, but we wouldn't worry about packaging and testing them properly.) Thanks! sage ___ ceph-users mailing list ceph-users@lists.ceph.commailto:ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] dropping old distros: el6, precise 12.04, debian wheezy?
Hi Jan, From my reading of Sage's email, hammer would continue to be supported on older distros, but new development would not target those releases. Was that your impression as well? As a former system administrator I feel your pain. Upgrading to new distros is a ton of work and incurs a ton of uncertainty and potential problems and liability. I truly think though that we will have a lot more luck testing and bug fixing new Ceph releases if we can focus specifically on new distro support and not get distracted with trying to maintain compatability for new Ceph releases on previous generation LTS distro releases. IE it's one thing if we've already tested say Hammer on those distros and minor bug fixes aren't likely to hit weird lurking kernel or distro bugs that aren't likely to get fixed. With new releases though, there's a ton of things we change, and some of them may be tied to expecting certain behavior in the kernel (Random example: XFS not blowing up with sparse writes when non-default extent sizes are used). At some point we need to stop make exceptions for stuff like this because an old kernel may not have a patch or behavior that we need to move Ceph forward. Mark On 07/30/2015 09:29 AM, Jan “Zviratko” Schermer wrote: I understand your reasons, but dropping support for LTS release like this is not right. You should /lege artis/ support every distribution the LTS release could have ever been installed on - that’s what the LTS label is for and what we rely on once we build a project on top of it CentOS 6 in particular is still very widely used and even installed, enterprise apps rely on it to this day. Someone out there is surely maintaining their LTS Ceph release on this distro and not having tested packages will hurt badly. We don’t want out project managers selecting EMC SAN over CEPH SDS because of such uncertainty, and you should benchmark yourself to those vendors, maybe... Every developer loves dropping support and concentrating on the bleeding edge interesting stuff but that’s not how it should work. Just my 2 cents... Jan On 30 Jul 2015, at 15:54, Sage Weil sw...@redhat.com mailto:sw...@redhat.com wrote: As time marches on it becomes increasingly difficult to maintain proper builds and packages for older distros. For example, as we make the systemd transition, maintaining the kludgey sysvinit and udev support for centos6/rhel6 is a pain in the butt and eats up time and energy to maintain and test that we could be spending doing more useful work. Dropping them would mean: - Ongoing development on master (and future versions like infernalis and jewel) would not be tested on these distros. - We would stop building upstream release packages on ceph.com http://ceph.com for new releases. - We would probably continue building hammer and firefly packages for future bugfix point releases. - The downstream distros would probably continue to package them, but the burden would be on them. For example, if Ubuntu wanted to ship Jewel on precise 12.04, they could, but they'd probably need to futz with the packaging and/or build environment to make it work. So... given that, I'd like to gauge user interest in these old distros. Specifically, CentOS6 / RHEL6 Ubuntu precise 12.04 Debian wheezy Would anyone miss them? In particular, dropping these three would mean we could drop sysvinit entirely and focus on systemd (and continue maintaining the existing upstart files for just a bit longer). That would be a relief. (The sysvinit files wouldn't go away in the source tree, but we wouldn't worry about packaging and testing them properly.) Thanks! sage ___ ceph-users mailing list ceph-users@lists.ceph.com mailto:ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
[ceph-users] Ceph Tech Talk Today!
Hey cephers, Just sending a friendly reminder that our online CephFS Tech Talk is happening today at 13:00 EDT (17:00 UTC). Please stop by and hear a technical deep dive on CephFS and ask any questions you might have. Thanks! http://ceph.com/ceph-tech-talks/ direct link to the video conference: https://bluejeans.com/172084437/browser -- Best Regards, Patrick McGarry Director Ceph Community || Red Hat http://ceph.com || http://community.redhat.com @scuttlemonkey || @ceph ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] dropping old distros: el6, precise 12.04, debian wheezy?
I understand your reasons, but dropping support for LTS release like this is not right. You should lege artis support every distribution the LTS release could have ever been installed on - that’s what the LTS label is for and what we rely on once we build a project on top of it CentOS 6 in particular is still very widely used and even installed, enterprise apps rely on it to this day. Someone out there is surely maintaining their LTS Ceph release on this distro and not having tested packages will hurt badly. We don’t want out project managers selecting EMC SAN over CEPH SDS because of such uncertainty, and you should benchmark yourself to those vendors, maybe... Every developer loves dropping support and concentrating on the bleeding edge interesting stuff but that’s not how it should work. Just my 2 cents... Jan On 30 Jul 2015, at 15:54, Sage Weil sw...@redhat.com wrote: As time marches on it becomes increasingly difficult to maintain proper builds and packages for older distros. For example, as we make the systemd transition, maintaining the kludgey sysvinit and udev support for centos6/rhel6 is a pain in the butt and eats up time and energy to maintain and test that we could be spending doing more useful work. Dropping them would mean: - Ongoing development on master (and future versions like infernalis and jewel) would not be tested on these distros. - We would stop building upstream release packages on ceph.com for new releases. - We would probably continue building hammer and firefly packages for future bugfix point releases. - The downstream distros would probably continue to package them, but the burden would be on them. For example, if Ubuntu wanted to ship Jewel on precise 12.04, they could, but they'd probably need to futz with the packaging and/or build environment to make it work. So... given that, I'd like to gauge user interest in these old distros. Specifically, CentOS6 / RHEL6 Ubuntu precise 12.04 Debian wheezy Would anyone miss them? In particular, dropping these three would mean we could drop sysvinit entirely and focus on systemd (and continue maintaining the existing upstart files for just a bit longer). That would be a relief. (The sysvinit files wouldn't go away in the source tree, but we wouldn't worry about packaging and testing them properly.) Thanks! sage ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] dropping old distros: el6, precise 12.04, debian wheezy?
It is possible I misunderstood Sage’s message - I apologize if that’s the case. This is what made me uncertain: - We would probably continue building hammer and firefly packages for future bugfix point releases. Decision for new releases (Infernalis, Jewel, K*) regarding distro support should be made officialy somewhere. Not sure if packages for them exist today and where? I don’t think new releases need to be made for CentOS 6 etc., parts of it are just too old for new stuff to make sense. But once that “commitment” is made (like http://ceph.com/docs/master/releases/ ) I expect them to be “supported for until the Ceph release itself is EOL. Of course whoever is “freeloading” must be ready for anything, anytime, no promises :-) Jan On 30 Jul 2015, at 16:47, Mark Nelson mnel...@redhat.com wrote: Hi Jan, From my reading of Sage's email, hammer would continue to be supported on older distros, but new development would not target those releases. Was that your impression as well? As a former system administrator I feel your pain. Upgrading to new distros is a ton of work and incurs a ton of uncertainty and potential problems and liability. I truly think though that we will have a lot more luck testing and bug fixing new Ceph releases if we can focus specifically on new distro support and not get distracted with trying to maintain compatability for new Ceph releases on previous generation LTS distro releases. IE it's one thing if we've already tested say Hammer on those distros and minor bug fixes aren't likely to hit weird lurking kernel or distro bugs that aren't likely to get fixed. With new releases though, there's a ton of things we change, and some of them may be tied to expecting certain behavior in the kernel (Random example: XFS not blowing up with sparse writes when non-default extent sizes are used). At some point we need to stop make exceptions for stuff like this because an old kernel may not have a patch or behavior that we need to move Ceph forward. Mark On 07/30/2015 09:29 AM, Jan “Zviratko” Schermer wrote: I understand your reasons, but dropping support for LTS release like this is not right. You should /lege artis/ support every distribution the LTS release could have ever been installed on - that’s what the LTS label is for and what we rely on once we build a project on top of it CentOS 6 in particular is still very widely used and even installed, enterprise apps rely on it to this day. Someone out there is surely maintaining their LTS Ceph release on this distro and not having tested packages will hurt badly. We don’t want out project managers selecting EMC SAN over CEPH SDS because of such uncertainty, and you should benchmark yourself to those vendors, maybe... Every developer loves dropping support and concentrating on the bleeding edge interesting stuff but that’s not how it should work. Just my 2 cents... Jan On 30 Jul 2015, at 15:54, Sage Weil sw...@redhat.com mailto:sw...@redhat.com wrote: As time marches on it becomes increasingly difficult to maintain proper builds and packages for older distros. For example, as we make the systemd transition, maintaining the kludgey sysvinit and udev support for centos6/rhel6 is a pain in the butt and eats up time and energy to maintain and test that we could be spending doing more useful work. Dropping them would mean: - Ongoing development on master (and future versions like infernalis and jewel) would not be tested on these distros. - We would stop building upstream release packages on ceph.com http://ceph.com for new releases. - We would probably continue building hammer and firefly packages for future bugfix point releases. - The downstream distros would probably continue to package them, but the burden would be on them. For example, if Ubuntu wanted to ship Jewel on precise 12.04, they could, but they'd probably need to futz with the packaging and/or build environment to make it work. So... given that, I'd like to gauge user interest in these old distros. Specifically, CentOS6 / RHEL6 Ubuntu precise 12.04 Debian wheezy Would anyone miss them? In particular, dropping these three would mean we could drop sysvinit entirely and focus on systemd (and continue maintaining the existing upstart files for just a bit longer). That would be a relief. (The sysvinit files wouldn't go away in the source tree, but we wouldn't worry about packaging and testing them properly.) Thanks! sage ___ ceph-users mailing list ceph-users@lists.ceph.com mailto:ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com ___ ceph-users mailing list
Re: [ceph-users] dropping old distros: el6, precise 12.04, debian wheezy?
-BEGIN PGP SIGNED MESSAGE- Hash: SHA256 I agree that for the distros and version in question, Ceph releases already released on them should provide bug support until EoL of Ceph or the distro version, whichever is shorter. Since we are so far into Infernalis and Jewel development cycle, would it be better to make Jewel the last supported release on these distros/versions? It would not be as much as a surprise and give admins a little more time to transition (we are currently looking to migrate to CentOS7 on the client side, but it will take some time so this is also a personal request). I can understand and appreciate that the Ceph team is stretched thin trying to maintain such a large breadth of compatibility. Thank you for reading, -BEGIN PGP SIGNATURE- Version: Mailvelope v0.13.1 Comment: https://www.mailvelope.com wsFcBAEBCAAQBQJVunuVCRDmVDuy+mK58QAAxvkP/0QqygaVYU9fV2bu3URb wvU4+2MT8W1zt9reYl+/42f709Us+7Abho90cMya61sIt+5t9h8oEms/5qwt ejx6CY3WX9x5Zobku1taUUtjtW3ZGZUcQ2JegO18WwqrSXfomd1vNvgB5JsJ KtDCv2/oCMLZWrFCXcAeoLkb/n3dwnctlsNs58911PERP7EAIMOVFBb5asee mUoyE06Nj5RUb5r2D7XfrcBxDb6eII0KLzpjbveGrUddoMgRPm2BVbab4N84 HDM330+3vg59pKsaRjU8wEgk4yuJtbxG1SaJMDEJKWHJNe+K5MFRX/FFtfMg 6P5ygDIn+7EK5EZLgRhJl0CTUCpXKa2/bc0tZzcwqV+YXmmAvAIHLa2hjpG9 1llSOlzKIR7lputbFbUHKJzgAOLG/nOOoUTGmnWqiTDEPrj3gOdI2sqaNWoX 50pT50pvkf64Pl1RXWPafveiX4pmZcK0jAYhqZTZ7Cpa0TfuyBfsDHZTzI6k YWunqhjqsm8mMFAr8ppjCK52e246iAdA1G7iXuPNE1yrtwhsHDc9VbqX8llT GaevOTQQygCZ/0b4adt+HWBQTZgLi0bxIyBEo7Pg75+f8yzawTpWWx5tM4JR yKkADT8+hid8PK6YwUrsaYnRhdRunAZv9BdvOulXyJ9azcCHSzlJWeYWHEwq krdm =m+W6 -END PGP SIGNATURE- Robert LeBlanc PGP Fingerprint 79A2 9CA4 6CC4 45DD A904 C70E E654 3BB2 FA62 B9F1 On Thu, Jul 30, 2015 at 7:54 AM, Sage Weil sw...@redhat.com wrote: As time marches on it becomes increasingly difficult to maintain proper builds and packages for older distros. For example, as we make the systemd transition, maintaining the kludgey sysvinit and udev support for centos6/rhel6 is a pain in the butt and eats up time and energy to maintain and test that we could be spending doing more useful work. Dropping them would mean: - Ongoing development on master (and future versions like infernalis and jewel) would not be tested on these distros. - We would stop building upstream release packages on ceph.com for new releases. - We would probably continue building hammer and firefly packages for future bugfix point releases. - The downstream distros would probably continue to package them, but the burden would be on them. For example, if Ubuntu wanted to ship Jewel on precise 12.04, they could, but they'd probably need to futz with the packaging and/or build environment to make it work. So... given that, I'd like to gauge user interest in these old distros. Specifically, CentOS6 / RHEL6 Ubuntu precise 12.04 Debian wheezy Would anyone miss them? In particular, dropping these three would mean we could drop sysvinit entirely and focus on systemd (and continue maintaining the existing upstart files for just a bit longer). That would be a relief. (The sysvinit files wouldn't go away in the source tree, but we wouldn't worry about packaging and testing them properly.) Thanks! sage ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] ceph-mon cpu usage
I saw this behavior when the servers are not in time sync. Check your ntp settings Dieter From: ceph-users ceph-users-boun...@lists.ceph.commailto:ceph-users-boun...@lists.ceph.com on behalf of Quentin Hartman qhart...@direwolfdigital.commailto:qhart...@direwolfdigital.com Date: Wednesday, July 29, 2015 at 5:47 PM To: Luis Periquito periqu...@gmail.commailto:periqu...@gmail.com Cc: Ceph Users ceph-users@lists.ceph.commailto:ceph-users@lists.ceph.com Subject: Re: [ceph-users] ceph-mon cpu usage I just had my ceph cluster exhibit this behavior (two of three mons eat all CPU, cluster becomes unusably slow) which is running 0.87.1 It seems to be tied to deep scrubbing, as the behavior almost immediately surfaces if that is turned on, but if it is off the behavior eventually seems to return to normal and stays that way while scrubbing is off. I have not yet found anything in the cluster to indicate a hardware problem. Any thoughts or further insights on this subject would be appreciated. QH On Sat, Jul 25, 2015 at 12:31 AM, Luis Periquito periqu...@gmail.commailto:periqu...@gmail.com wrote: I think I figured out! All 4 of the OSDs on one host (OSD 107-110) were sending massive amounts of auth requests to the monitors, seeming to overwhelm them. Weird bit is that I removed them (osd crush remove, auth del, osd rm), dd the box and all of the disks, reinstalled and guess what? They are still doing a lot of requests to the MONs... this will require some further investigations. As this is happening during my holidays, I just disabled them, and will investigate further when I get back. On Fri, Jul 24, 2015 at 11:11 PM, Kjetil Jørgensen kje...@medallia.commailto:kje...@medallia.com wrote: It sounds slightly similar to what I just experienced. I had one monitor out of three, which seemed to essentially run one core at full tilt continuously, and had it's virtual address space allocated at the point where top started calling it Tb. Requests hitting this monitor did not get very timely responses (although; I don't know if this were happening consistently or arbitrarily). I ended up re-building the monitor from the two healthy ones I had, which made the problem go away for me. After the fact inspection of the monitor I ripped out, clocked it in at 1.3Gb compared to the 250Mb of the other two, after rebuild they're all comparable in size. In my case; this started out for me on firefly, and persisted after upgrading to hammer. Which prompted the rebuild, suspecting that in my case it were related to something persistent for this monitor. I do not have that much more useful to contribute to this discussion, since I've more-or-less destroyed any evidence by re-building the monitor. Cheers, KJ On Fri, Jul 24, 2015 at 1:55 PM, Luis Periquito periqu...@gmail.commailto:periqu...@gmail.com wrote: The leveldb is smallish: around 70mb. I ran debug mon = 10 for a while, but couldn't find any interesting information. I would run out of space quite quickly though as the log partition only has 10g. On 24 Jul 2015 21:13, Mark Nelson mnel...@redhat.commailto:mnel...@redhat.com wrote: On 07/24/2015 02:31 PM, Luis Periquito wrote: Now it's official, I have a weird one! Restarted one of the ceph-mons with jemalloc and it didn't make any difference. It's still using a lot of cpu and still not freeing up memory... The issue is that the cluster almost stops responding to requests, and if I restart the primary mon (that had almost no memory usage nor cpu) the cluster goes back to its merry way responding to requests. Does anyone have any idea what may be going on? The worst bit is that I have several clusters just like this (well they are smaller), and as we do everything with puppet, they should be very similar... and all the other clusters are just working fine, without any issues whatsoever... We've seen cases where leveldb can't compact fast enough and memory balloons, but it's usually associated with extreme CPU usage as well. It would be showing up in perf though if that were the case... On 24 Jul 2015 10:11, Jan Schermer j...@schermer.czmailto:j...@schermer.cz mailto:j...@schermer.czmailto:j...@schermer.cz wrote: You don’t (shouldn’t) need to rebuild the binary to use jemalloc. It should be possible to do something like LD_PRELOAD=/usr/lib/x86_64-linux-gnu/libjemalloc.so.1 ceph-osd … The last time we tried it segfaulted after a few minutes, so YMMV and be careful. Jan On 23 Jul 2015, at 18:18, Luis Periquito periqu...@gmail.commailto:periqu...@gmail.com mailto:periqu...@gmail.commailto:periqu...@gmail.com wrote: Hi Greg, I've been looking at the tcmalloc issues, but did seem to affect osd's, and I do notice it in heavy read workloads (even after the patch and increasing TCMALLOC_MAX_TOTAL_THREAD_CACHE_BYTES=134217728). This is affecting the mon process though. looking at perf top I'm getting most of the CPU usage
Re: [ceph-users] ceph-mon cpu usage
Thanks for the suggestion. NTP is fine in my case. Turns out it was a networking problem that wasn't triggering error counters on the NICs so it took a bit to track it down. QH On Thu, Jul 30, 2015 at 4:16 PM, Spillmann, Dieter dieter.spillm...@arris.com wrote: I saw this behavior when the servers are not in time sync. Check your ntp settings Dieter From: ceph-users ceph-users-boun...@lists.ceph.com on behalf of Quentin Hartman qhart...@direwolfdigital.com Date: Wednesday, July 29, 2015 at 5:47 PM To: Luis Periquito periqu...@gmail.com Cc: Ceph Users ceph-users@lists.ceph.com Subject: Re: [ceph-users] ceph-mon cpu usage I just had my ceph cluster exhibit this behavior (two of three mons eat all CPU, cluster becomes unusably slow) which is running 0.87.1 It seems to be tied to deep scrubbing, as the behavior almost immediately surfaces if that is turned on, but if it is off the behavior eventually seems to return to normal and stays that way while scrubbing is off. I have not yet found anything in the cluster to indicate a hardware problem. Any thoughts or further insights on this subject would be appreciated. QH On Sat, Jul 25, 2015 at 12:31 AM, Luis Periquito periqu...@gmail.com wrote: I think I figured out! All 4 of the OSDs on one host (OSD 107-110) were sending massive amounts of auth requests to the monitors, seeming to overwhelm them. Weird bit is that I removed them (osd crush remove, auth del, osd rm), dd the box and all of the disks, reinstalled and guess what? They are still doing a lot of requests to the MONs... this will require some further investigations. As this is happening during my holidays, I just disabled them, and will investigate further when I get back. On Fri, Jul 24, 2015 at 11:11 PM, Kjetil Jørgensen kje...@medallia.com wrote: It sounds slightly similar to what I just experienced. I had one monitor out of three, which seemed to essentially run one core at full tilt continuously, and had it's virtual address space allocated at the point where top started calling it Tb. Requests hitting this monitor did not get very timely responses (although; I don't know if this were happening consistently or arbitrarily). I ended up re-building the monitor from the two healthy ones I had, which made the problem go away for me. After the fact inspection of the monitor I ripped out, clocked it in at 1.3Gb compared to the 250Mb of the other two, after rebuild they're all comparable in size. In my case; this started out for me on firefly, and persisted after upgrading to hammer. Which prompted the rebuild, suspecting that in my case it were related to something persistent for this monitor. I do not have that much more useful to contribute to this discussion, since I've more-or-less destroyed any evidence by re-building the monitor. Cheers, KJ On Fri, Jul 24, 2015 at 1:55 PM, Luis Periquito periqu...@gmail.com wrote: The leveldb is smallish: around 70mb. I ran debug mon = 10 for a while, but couldn't find any interesting information. I would run out of space quite quickly though as the log partition only has 10g. On 24 Jul 2015 21:13, Mark Nelson mnel...@redhat.com wrote: On 07/24/2015 02:31 PM, Luis Periquito wrote: Now it's official, I have a weird one! Restarted one of the ceph-mons with jemalloc and it didn't make any difference. It's still using a lot of cpu and still not freeing up memory... The issue is that the cluster almost stops responding to requests, and if I restart the primary mon (that had almost no memory usage nor cpu) the cluster goes back to its merry way responding to requests. Does anyone have any idea what may be going on? The worst bit is that I have several clusters just like this (well they are smaller), and as we do everything with puppet, they should be very similar... and all the other clusters are just working fine, without any issues whatsoever... We've seen cases where leveldb can't compact fast enough and memory balloons, but it's usually associated with extreme CPU usage as well. It would be showing up in perf though if that were the case... On 24 Jul 2015 10:11, Jan Schermer j...@schermer.cz mailto:j...@schermer.cz wrote: You don’t (shouldn’t) need to rebuild the binary to use jemalloc. It should be possible to do something like LD_PRELOAD=/usr/lib/x86_64-linux-gnu/libjemalloc.so.1 ceph-osd … The last time we tried it segfaulted after a few minutes, so YMMV and be careful. Jan On 23 Jul 2015, at 18:18, Luis Periquito periqu...@gmail.com mailto:periqu...@gmail.com wrote: Hi Greg, I've been looking at the tcmalloc issues, but did seem to affect osd's, and I do notice it in heavy read workloads (even after the patch and increasing TCMALLOC_MAX_TOTAL_THREAD_CACHE_BYTES=134217728). This is affecting the mon process though. looking at perf top I'm getting most of the CPU usage in
Re: [ceph-users] which kernel version can help avoid kernel client deadlock
Date: Thu, 30 Jul 2015 13:11:11 +0300 Subject: Re: [ceph-users] which kernel version can help avoid kernel client deadlock From: idryo...@gmail.com To: zhangz.da...@outlook.com CC: chaofa...@owtware.com; ceph-users@lists.ceph.com On Thu, Jul 30, 2015 at 12:46 PM, Z Zhang zhangz.da...@outlook.com wrote: Date: Thu, 30 Jul 2015 11:37:37 +0300 Subject: Re: [ceph-users] which kernel version can help avoid kernel client deadlock From: idryo...@gmail.com To: zhangz.da...@outlook.com CC: chaofa...@owtware.com; ceph-users@lists.ceph.com On Thu, Jul 30, 2015 at 10:29 AM, Z Zhang zhangz.da...@outlook.com wrote: Subject: Re: [ceph-users] which kernel version can help avoid kernel client deadlock From: chaofa...@owtware.com Date: Thu, 30 Jul 2015 13:16:16 +0800 CC: idryo...@gmail.com; ceph-users@lists.ceph.com To: zhangz.da...@outlook.com On Jul 30, 2015, at 12:48 PM, Z Zhang zhangz.da...@outlook.com wrote: We also hit the similar issue from time to time on centos with 3.10.x kernel. By iostat, we can see kernel rbd client's util is 100%, but no r/w io, and we can't umount/unmap this rbd client. After restarting OSDs, it will become normal. 3.10.x is rather vague, what is the exact version you saw this on? Can you provide syslog logs (I'm interested in dmesg)? The kernel version should be 3.10.0. I don't have sys logs at hand. It is not easily reproduced, and it happened at very low memory situation. We are running DB instances over rbd as storage. DB instances will use lot of memory when running high concurrent rw, and after running for a long time, rbd might hit this problem, but not always. Enabling rbd log makes our system behave strange during our test. I back-ported one of your fixes: https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/drivers/block/rbd.c?id=5a60e87603c4c533492c515b7f62578189b03c9c So far test looks fine for few days, but still under observation. So want to know if there are some other fixes? I'd suggest following 3.10 stable series (currently at 3.10.84). The fix you backported is crucial in low memory situations, so I wouldn't be surprised if it alone fixed your problem. (It is not in 3.10.84, I assume it'll show up in 3.10.85 - for now just apply your backport.) cool, looking forward 3.10.85 to see what else would be brought in. Thanks. Thanks, Ilya ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
[ceph-users] RGW + civetweb + SSL
Hello, I’d like to know if someone know how to setup a SSL implementation of RGW with civetweb? The only “documentation” that I found about that is a “bug” - http://tracker.ceph.com/issues/11239 - which I’d like to know if this kind of implementation really works? Regards. Italo Santos http://italosantos.com.br/ ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] questions on editing crushmap for ceph cache tier
On Jul 31, 2015, at 2:55 AM, Robert LeBlanc rob...@leblancnet.us wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA256 You are close... I've done it by creating a new SSD root in the CRUSH map, then put the SSD OSDs into a -ssd entry. I then created a new crush rule to choose from the SSD root, then have the tiering pool use that rule. If you look at the example in the document and think of ceph-osd-ssd-server-1 and ceph-osd-platter-server-1 as the same physical server with just logical separation, you can follow the rest of the example. You will need to modify either the ceph.conf to have the ssd OSDs have a different CRUSH map location or write a location script to do it automatically [1]. ceph.conf example: [osd.70] crush location = root=ssd host=nodez-ssd Thanks. It works. Modifying ceph.conf is very important. Otherwise, after I start the OSDs, it will automatically go back to original host. To do it programmaticly, you can use /usr/bin/ceph-crush-location as a base. I extended it by finding the device, then searching through hdparam to see if the rotation was greater than zero. If it wasn't then I output the hostname with the ssd portion, otherwise just the hostname. It was only a few lines of code, but I can't find it right now. I was waiting for ACLs to be modified so that I could query the data center location from our inventory management system (that was a few months ago) and I'm still waiting. [1] http://lists.ceph.com/pipermail/ceph-users-ceph.com/2015-January/045673.html -BEGIN PGP SIGNATURE- Version: Mailvelope v0.13.1 Comment: https://www.mailvelope.com wsFcBAEBCAAQBQJVunMXCRDmVDuy+mK58QAAq04P+QH95haU34fWZ5PsPIxv oY7JReywHwP3mWBO2XkaIg/l4AYV/HCBckBTSyp+GGFAtMeVEndiHCYTUf+F kQJjIk1jZoN+WTLnD8nsfDMrmVmforyGcG4Y399C4cCkBmeoU3jeGeKx+Unx dxiW6flRH5GPCazQdAAbXgb8InynUZ/EqmTY0FCDuLQ3CEXELuM8IReKwz0X 9LYqVIV+tdE1Ff2nDnLQlYYVpVv5K0y4TXBj8JzYH/41XbEws2GQnhb6b8zW aopUDC9RsNGtzWf4RDg8X3LDHrw7IBtAuJf+PHbcq3Y4cmPf5Z0TYiS1bqn1 19kj3EhDVVdW1KUue2S3GemyP0+bIypA/VDGzFXgv9g5oKN0bXPOjuFKAD2Q 7Rc2yoW70LACgL0a2KiRPRt8e6Jz5/vG6GijZvxTTZfkDPKHBOPPA3mFyAS4 FGmu39/q5VP7V+CepaKjbGNUWRzlzOOcc4ybk3dmktYEFOTw4QZLczBGJ8s4 I+AdYDjiQOAG3n3xixqRFOb4URjfKOrUbnHfNVQJU+qfYfV1RBLThhRjiv0v O+oiKiWuugZicHniTfHuOYePgxbs9eU2Hk8VRVk9ievXuRynDrH7D+IeUzUJ JGoj01YM60Ul1XJPWatMoM+435hcHrGd0rJ3bi91DOrZmT55X4jjdUA8z/3Y xMZE =Tqaw -END PGP SIGNATURE- Robert LeBlanc PGP Fingerprint 79A2 9CA4 6CC4 45DD A904 C70E E654 3BB2 FA62 B9F1 On Wed, Jul 29, 2015 at 9:21 PM, van chaofa...@owtware.com wrote: Hi, list, Ceph cache tier seems very promising for performance. According to http://ceph.com/docs/master/rados/operations/crush-map/#placing-different-pools-on-different-osds , I need to create a new pool based on SSD OSDs。 Currently, I’ve two servers with several HDD-based OSDs. I mean to add one SSD-based OSD for each server, and then use these two OSDs to build a cache pool. But I’ve found problems editing crushmap. The example in the link use two new hosts to build SSD OSDs and then create a new ruleset take the new hosts. But in my environment, I do not have new servers to use. Can I create a ruleset choose part of OSDs in a host? For example, as the crushmap shown below, osd.2 and osd.5 are new added SSD-based OSDs, how can I create a ruleset choose these two OSDs only and how can I avoid default ruleset to choose osd.2 and osd.5? Is this possible, or I have to add a new server to deploy cache tier? Thanks. host node0 { id -2 alg straw hash 0 item osd.0 weight 1.0 # HDD item osd.1 weight 1.0 # HDD item osd.2 weight 0.5 # SSD } host node1 { id -3 alg straw hash 0 item osd.3 weight 1.0 # HDD item osd.4 weight 1.0 # HDD item osd.5 weight 0.5 # SSD } root default { id -1 # do not change unnecessarily # weight 1.560 alg straw hash 0 # rjenkins1 item node0 weight 2.5 item node1 weight 2.5 } # typical ruleset rule replicated_ruleset { ruleset 0 type replicated min_size 1 max_size 10 step take default step chooseleaf firstn 0 type host step emit } van chaofa...@owtware.com ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] Check networking first?
wouldn't it be nice that ceph does something like this in background (some sort of network-scrub). debugging network like this is not that easy (can't expect admins to install e.g. perfsonar on all nodes and/or clients) something like: every X min, each service X pick a service Y on another host (assuming X and Y will exchange some communication at some point; like osd with other osd), send 1MB of data, and make the timing data available so we can monitor it and detect underperforming links over time. ideally clients also do this, but not sure where they should report/store the data. interpreting the data can be a bit tricky, but extreme outliers will be spotted easily, and the main issue with this sort of debugging is collecting the data. simply reporting / keeping track of ongoing communications is already a big step forward, but then we need to have the size of the exchanged data to allow interpretation (and the timing should be about the network part, not e.g. flush data to disk in case of an osd). (and obviously sampling is enough, no need to have details of every bit send). stijn On 07/30/2015 08:04 PM, Mark Nelson wrote: Thanks for posting this! We see issues like this more often than you'd think. It's really important too because if you don't figure it out the natural inclination is to blame Ceph! :) Mark On 07/30/2015 12:50 PM, Quentin Hartman wrote: Just wanted to drop a note to the group that I had my cluster go sideways yesterday, and the root of the problem was networking again. Using iperf I discovered that one of my nodes was only moving data at 1.7Mb / s. Moving that node to a different switch port with a different cable has resolved the problem. It took awhile to track down because none of the server-side error metrics for disk or network showed anything was amiss, and I didn't think to test network performance (as suggested in another thread) until well into the process. Check networking first! QH ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] fuse mount in fstab
Hi More info about this issue, we have opened a ticket to redhat here is the feedback: https://bugzilla.redhat.com/show_bug.cgi?id=1248003 Cheers Alvaro On 16/07/15 15:19, Alvaro Simon Garcia wrote: Hi I have tested a bit this with different ceph-fuse versions and linux distros and it seems a mount issue in CentOS7. The problem is that mount tries to find first the param=value from fstab fs_spec field into the blkid block device attributes and of course this flag is not there and you always get an error like this: || mount: can't find param=value and stops here, the mount values are never parsed by /sbin/mount.fuse.ceph helper... The only workaround that I found without change mount version is to change the spurious = by another special character like a colon for example: id:admin /mnt/ceph fuse.ceph defaults 0 0 but you also have to change /sbin/mount.fuse.ceph parser: ... # convert device string to options fs_spec=`echo $1 | sed 's/:/=/g'` cephargs='--'`echo $fs_spec | sed 's/,/ --/g'` ... but this is a bit annoying... someone else has found the same mount fuse issue in RHEL7 or CentOS? Cheers Alvaro On 09/07/15 12:22, Kenneth Waegeman wrote: Hmm, it looks like a version issue.. I am testing with these versions on centos7: ~]# mount -V mount from util-linux 2.23.2 (libmount 2.23.0: selinux, debug, assert) ~]# ceph-fuse -v ceph version 0.94.1 (e4bfad3a3c51054df7e537a724c8d0bf9be972ff) This do not work.. On my fedora box, with these versions from repo: # mount -V mount from util-linux 2.24.2 (libmount 2.24.0: selinux, debug, assert) # ceph-fuse -v ceph version 0.80.9 (b5a67f0e1d15385bc0d60a6da6e7fc810bde6047) this works.. Which versions are you running? And does someone knows from which versions , or which version combinations do work? Thanks a lot! K On 07/09/2015 11:53 AM, Thomas Lemarchand wrote: Hello Kenneth, I have a working ceph fuse in fstab. Only difference I see it that I don't use conf, your configuration file is at the default path anyway. I tried it with and without conf, but it always complains about id id=recette-files-rw,client_mountpoint=/recette-files/files /mnt/wimi/ceph-files fuse.ceph noatime,_netdev 0 0 ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] rbd-fuse Transport endpoint is not connected
On Wed, Jul 29, 2015 at 11:42 PM, pixelfairy pixelfa...@gmail.com wrote: copied ceph.conf from the servers. hope this helps. should this be concidered an unsupported feature? # rbd-fuse /cmnt -c /etc/ceph/ceph.conf -d FUSE library version: 2.9.2 nullpath_ok: 0 nopath: 0 utime_omit_ok: 0 unique: 1, opcode: INIT (26), nodeid: 0, insize: 56, pid: 0 INIT: 7.22 flags=0xf7fb max_readahead=0x0002 Error connecting to cluster: No such file or directory This is an error from rados_connect(), so I'd say there is something wrong with your ceph.conf or the environment. We can try to debug this, but yeah, rbd-fuse, as it is, is not intended for serious use. Thanks, Ilya ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
[ceph-users] Squeeze packages for 0.94.2
Hello, it seems that there are no Debian Squeeze packages in the repository for the current Hammer version. Is this an oversight or is there another reason those are not provided? Sebastian ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] How to identify MDS client failing to respond to capability release?
For sufficiently recent clients we do this for you (clients send some metadata like hostname, which is used in the MDS to generate an easier-to-understand identifier). To do it by hand, use the admin socket command ceph daemon mds.id session ls command, and look out for the client IP addresses in the address part of each entry. By the way, there have been various other bugs in 3.13-ish kernels, so you should consider using either a more recent kernel or a fuse client. John On 30/07/15 08:32, Oliver Schulz wrote: Hello Ceph Experts, lately, ceph status on our cluster often states: mds0: Client CLIENT_ID failing to respond to capability release How can I identify which client is at fault (hostname or IP address) from the CLIENT_ID? What could be the source of the failing to respond to capability release - Linux kernel on the client too old? We use ceph-0.94.2 on the cluster, and the CephFS kernel client on the clients (kernel 3.13.0 on Ubuntu 12.04 and Ubuntu 14.04). But it's possible that there's a machine with an older Kernel around somewhere ... Cheers and thanks, Oliver ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
[ceph-users] mount rbd image with iscsi
hi, I am trying to mount an RBD image using iSCSI following URL : *http://www.sebastien-han.fr/blog/2014/07/07/start-with-the-rbd-support-for-tgt/ http://www.sebastien-han.fr/blog/2014/07/07/start-with-the-rbd-support-for-tgt/* However, I don't get rbd flag when I give the command sudo tgtadm --lld iscsi --op show --mode system | grep rbd rbd (bsoflags sync:direct) *--- this is not showing* Can someone help with this so that I can mount RBD image to iSCSI thanks.. ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] Unable to mount Format 2 striped RBD image
hi Ilya, I had used the below command to create the rbd image rbd -p fool create strpimg --image-format 2 --order 22 --size 2048M --stripe-unit 65536 --stripe-count 3 --image-feature striping --image-shared I am confused when you say why I use sysfs instead of rbd cli tool. Can you please help me? I dont wish to use qemu or virtmanager. I wish to use this as block device on client PC Thanks. Daleep On Wed, Jul 29, 2015 at 4:19 PM, Ilya Dryomov idryo...@gmail.com wrote: On Wed, Jul 29, 2015 at 1:45 PM, Daleep Bais dal...@mask365.com wrote: Hi, I have created a format 2 striped image, however, I am not able to mount it on client machine.. # rbd -p foo info strpimg rbd image 'strpimg': size 2048 MB in 513 objects order 22 (4096 kB objects) block_name_prefix: rbd_data.20c942ae8944a format: 2 features: striping flags: stripe unit: 65536 bytes stripe count: 3 Getting below error when trying to mount using echo --- /sys/bus/rbd/add write error : invalid arguement if i use a format 1 image, I am able to mount the RBD to client and use it. Custom striping settings (i.e. non-default stripe_unit and stripe_count) are not yet supported by the kernel client. Unrelated, why are you using sysfs directly instead of rbd cli tool? Thanks, Ilya ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] which kernel version can help avoid kernel client deadlock
On Thu, Jul 30, 2015 at 10:29 AM, Z Zhang zhangz.da...@outlook.com wrote: Subject: Re: [ceph-users] which kernel version can help avoid kernel client deadlock From: chaofa...@owtware.com Date: Thu, 30 Jul 2015 13:16:16 +0800 CC: idryo...@gmail.com; ceph-users@lists.ceph.com To: zhangz.da...@outlook.com On Jul 30, 2015, at 12:48 PM, Z Zhang zhangz.da...@outlook.com wrote: We also hit the similar issue from time to time on centos with 3.10.x kernel. By iostat, we can see kernel rbd client's util is 100%, but no r/w io, and we can't umount/unmap this rbd client. After restarting OSDs, it will become normal. 3.10.x is rather vague, what is the exact version you saw this on? Can you provide syslog logs (I'm interested in dmesg)? Thanks, Ilya ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] Unable to mount Format 2 striped RBD image
On Thu, Jul 30, 2015 at 11:15 AM, Daleep Bais dal...@mask365.com wrote: hi Ilya, I had used the below command to create the rbd image rbd -p fool create strpimg --image-format 2 --order 22 --size 2048M --stripe-unit 65536 --stripe-count 3 --image-feature striping --image-shared I am confused when you say why I use sysfs instead of rbd cli tool. Well, there is no way to create an image via sysfs, but you can map an image by writing to /sys/bus/rbd/... instead of using rbd map. You wrote this: Getting below error when trying to mount using echo --- /sys/bus/rbd/add write error : invalid arguement which is why I asked if you had used sysfs directly, instead of something like rbd map strpimg. Can you please help me? I dont wish to use qemu or virtmanager. I wish to use this as block device on client PC You'd have to drop --stripe-unit 65536 --stripe-count 3 from your rbd create for that to work, at least for now. Thanks, Ilya ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] Recovery question
For the record, I have been able to recover. Thank you very much for the guidance. I hate searching the web and finding only partial information on threads like this, so I'm going to document and post what I've learned as best I can in hopes that it will help someone else out in the future. -- Peter Hinman On 7/29/2015 5:15 PM, Robert LeBlanc wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA256 If you had multiple monitors, you should recover if possible more than 50% of them (they will need to form a quorum). If you can't, it is messy but, you can manually remove enough monitors to start a quorum. From /etc/ceph/ you will want the keyring and the ceph.conf at a minimim. The keys for the monitor I think are in the store.db which will let the monitors start, but the keyring has the admin key which lets you manage the cluster once you get it up. rbdmap is not needed for recovery (only automatically mounting RBDs at boot time), we can deal with that later if needed. - Robert LeBlanc PGP Fingerprint 79A2 9CA4 6CC4 45DD A904 C70E E654 3BB2 FA62 B9F1 On Wed, Jul 29, 2015 at 4:40 PM, Peter Hinman wrote: Ok - that is encouraging. I've believe I've got data from a previous monitor. I see files in a store.db dated yesterday, with a MANIFEST- file that is significantly greater than the MANIFEST-07 file listed for the current monitors. I've actually found data for two previous monitors. Any idea which one I should select? The one with the highest manifest number? The most recent time stamp? What files should I be looking for in /etc/conf? Just the keyring and rbdmap files? How important is it to use the same keyring file? -- Peter Hinman On 7/29/2015 3:47 PM, Robert LeBlanc wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA256 The default is /var/lib/ceph/mon/- (/var/lib/ceph/mon/ceph-mon1 for me). You will also need the information from /etc/ceph/ to reconstruct the data. I *think* you should be able to just copy this to a new box with the same name and IP address and start it up. I haven't actually done this, so there still may be some bumps. - Robert LeBlanc PGP Fingerprint 79A2 9CA4 6CC4 45DD A904 C70E E654 3BB2 FA62 B9F1 On Wed, Jul 29, 2015 at 3:44 PM, Peter Hinman wrote: Thanks Robert - Where would that monitor data (database) be found? -- Peter Hinman On 7/29/2015 3:39 PM, Robert LeBlanc wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA256 If you built new monitors, this will not work. You would have to recover the monitor data (database) from at least one monitor and rebuild the monitor. The new monitors would not have any information about pools, OSDs, PGs, etc to allow an OSD to be rejoined. - Robert LeBlanc PGP Fingerprint 79A2 9CA4 6CC4 45DD A904 C70E E654 3BB2 FA62 B9F1 On Wed, Jul 29, 2015 at 2:46 PM, Peter Hinman wrote: Hi Greg - So at the moment, I seem to be trying to resolve a permission error. === osd.3 === Mounting xfs on stor-2:/var/lib/ceph/osd/ceph-3 2015-07-29 13:35:08.809536 7f0a0262e700 0 librados: osd.3 authentication error (1) Operation not permitted Error connecting to cluster: PermissionError failed: 'timeout 30 /usr/bin/ceph -c /etc/ceph/ceph.conf --name=osd.3 --keyring=/var/lib/ceph/osd/ceph-3/keyring osd crush create-or-move -- 3 3.64 host=stor-2 root=default' ceph-disk: Error: ceph osd start failed: Command '['/usr/sbin/service', 'ceph', '--cluster', 'ceph', 'start', 'osd.3']' returned non-zero exit status 1 ceph-disk: Error: One or more partitions failed to activate Is there a way to identify the cause of this PermissionError? I've copied the client.bootstrap-osd key from the output of ceph auth list, and pasted it into /var/lib/ceph/bootstrap-osd/ceph.keyring, but that has not resolve the error. But it sounds like you are saying that even once I get this resolved, I have no hope of recovering the data? -- Peter Hinman On 7/29/2015 1:57 PM, Gregory Farnum wrote: This sounds like you're trying to reconstruct a cluster after destroying the monitors. That is...not going to work well. The monitors define the cluster and you can't move OSDs into different clusters. We have ideas for how to reconstruct monitors and it can be done manually with a lot of hassle, but the process isn't written down and there aren't really fools I help with it. :/ -Greg On Wed, Jul 29, 2015 at 5:48 PM Peter Hinman wrote: I've got a situation that seems on the surface like it should be recoverable, but I'm struggling to understand how to do it. I had a cluster of 3 monitors, 3 osd disks, and 3 journal ssds. After multiple hardware failures, I pulled the 3 osd disks and 3 journal ssds and am attempting to bring them back up again on new hardware in a new cluster. I see plenty of documentation on how to zap and initialize and add new osds, but I don't see anything on rebuilding with existing osd disks. Could somebody provide guidance on how to do this? I'm
[ceph-users] ceph osd mounting issue with ocfs2
Hi All, We are using ceph with two OSD and three clients. Clients try to mount with OCFS2 file system. Here when i start mounting only two clients i can able to mount properly and third client giving below errors. Some time i can able to mount third client but data not sync to third client. mount /dev/rbd/rbd/integdownloads /soho/build/downloads mount.ocfs2: Invalid argument while mounting /dev/rbd0 on /soho/build/downloads. Check 'dmesg' for more information on this error. dmesg [1280548.676688] (mount.ocfs2,1807,4):dlm_send_nodeinfo:1294 ERROR: node mismatch -22, node 0 [1280548.676766] (mount.ocfs2,1807,4):dlm_try_to_join_domain:1681 ERROR: status = -22 [1280548.677278] (mount.ocfs2,1807,8):dlm_join_domain:1950 ERROR: status = -22 [1280548.677443] (mount.ocfs2,1807,8):dlm_register_domain:2210 ERROR: status = -22 [1280548.677541] (mount.ocfs2,1807,8):o2cb_cluster_connect:368 ERROR: status = -22 [1280548.677602] (mount.ocfs2,1807,8):ocfs2_dlm_init:2988 ERROR: status = -22 [1280548.677703] (mount.ocfs2,1807,8):ocfs2_mount_volume:1864 ERROR: status = -22 [1280548.677800] ocfs2: Unmounting device (252,0) on (node 0) [1280548.677808] (mount.ocfs2,1807,8):ocfs2_fill_super:1238 ERROR: status = -22 OCFS2 configuration cluster: node_count=3 heartbeat_mode = local name=ocfs2 node: ip_port = ip_address = 192.168.112.192 number = 0 name = integ-hm5 cluster = ocfs2 node: ip_port = ip_address = 192.168.113.42 number = 1 name = integ-soho cluster = ocfs2 node: ip_port = 7778 ip_address = 192.168.112.115 number = 2 name = integ-hm2 cluster = ocfs2 Ceph configuration # ceph -s cluster 944fa0af-b7be-45a9-93ff-b9907cfaee3f health HEALTH_OK monmap e2: 3 mons at {integ-hm5=192.168.112.192:6789/0,integ-hm6=192.168.112.193:6789/0,integ-hm7=192.168.112.194:6789/0} election epoch 54, quorum 0,1,2 integ-hm5,integ-hm6,integ-hm7 osdmap e10: 2 osds: 2 up, 2 in pgmap v32626: 64 pgs, 1 pools, 10293 MB data, 8689 objects 14575 MB used, 23651 GB / 24921 GB avail 64 active+clean client io 2047 B/s rd, 1023 B/s wr, 2 op/s Regards GJ ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
[ceph-users] How to identify MDS client failing to respond to capability release?
Hello Ceph Experts, lately, ceph status on our cluster often states: mds0: Client CLIENT_ID failing to respond to capability release How can I identify which client is at fault (hostname or IP address) from the CLIENT_ID? What could be the source of the failing to respond to capability release - Linux kernel on the client too old? We use ceph-0.94.2 on the cluster, and the CephFS kernel client on the clients (kernel 3.13.0 on Ubuntu 12.04 and Ubuntu 14.04). But it's possible that there's a machine with an older Kernel around somewhere ... Cheers and thanks, Oliver ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] dropping old distros: el6, precise 12.04, debian wheezy?
Sage Weil sw...@redhat.com 2015-07-30 06:54: As time marches on it becomes increasingly difficult to maintain proper builds and packages for older distros. For example, as we make the systemd transition, maintaining the kludgey sysvinit and udev support for centos6/rhel6 is a pain in the butt and eats up time and energy to maintain and test that we could be spending doing more useful work. Dropping them would mean: - Ongoing development on master (and future versions like infernalis and jewel) would not be tested on these distros. - We would stop building upstream release packages on ceph.com for new releases. - We would probably continue building hammer and firefly packages for future bugfix point releases. - The downstream distros would probably continue to package them, but the burden would be on them. For example, if Ubuntu wanted to ship Jewel on precise 12.04, they could, but they'd probably need to futz with the packaging and/or build environment to make it work. So... given that, I'd like to gauge user interest in these old distros. Specifically, CentOS6 / RHEL6 Ubuntu precise 12.04 Debian wheezy Would anyone miss them? In particular, dropping these three would mean we could drop sysvinit entirely and focus on systemd (and continue maintaining the existing upstart files for just a bit longer). That would be a relief. (The sysvinit files wouldn't go away in the source tree, but we wouldn't worry about packaging and testing them properly.) Thanks! sage As I still haven't heard or seen about any upstream distros for Debian Jessie (see also [1]), I am still running Debian Wheezy and as that is supposed to be supported for another ~4 years by Debian, it would be very nice if there were at least stability and security fixes backported for the upstream ceph package repositories for that platform. Additionally, I'll note that I'm personally likely to continue to use sysvinit so long as I still can, even when I am able to make the switch to Jessie. Thanks, Brian [1] http://www.spinics.net/lists/ceph-users/msg19959.html signature.asc Description: Digital signature ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] dropping old distros: el6, precise 12.04, debian wheezy?
Hi, dropping debian wheezy are quite fast - till now there aren't packages for jessie?! Dropping of squeeze I understand, but wheezy at this time? Udo On 30.07.2015 15:54, Sage Weil wrote: As time marches on it becomes increasingly difficult to maintain proper builds and packages for older distros. For example, as we make the systemd transition, maintaining the kludgey sysvinit and udev support for centos6/rhel6 is a pain in the butt and eats up time and energy to maintain and test that we could be spending doing more useful work. Dropping them would mean: - Ongoing development on master (and future versions like infernalis and jewel) would not be tested on these distros. - We would stop building upstream release packages on ceph.com for new releases. - We would probably continue building hammer and firefly packages for future bugfix point releases. - The downstream distros would probably continue to package them, but the burden would be on them. For example, if Ubuntu wanted to ship Jewel on precise 12.04, they could, but they'd probably need to futz with the packaging and/or build environment to make it work. So... given that, I'd like to gauge user interest in these old distros. Specifically, CentOS6 / RHEL6 Ubuntu precise 12.04 Debian wheezy Would anyone miss them? In particular, dropping these three would mean we could drop sysvinit entirely and focus on systemd (and continue maintaining the existing upstart files for just a bit longer). That would be a relief. (The sysvinit files wouldn't go away in the source tree, but we wouldn't worry about packaging and testing them properly.) Thanks! sage ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
[ceph-users] Check networking first?
Just wanted to drop a note to the group that I had my cluster go sideways yesterday, and the root of the problem was networking again. Using iperf I discovered that one of my nodes was only moving data at 1.7Mb / s. Moving that node to a different switch port with a different cable has resolved the problem. It took awhile to track down because none of the server-side error metrics for disk or network showed anything was amiss, and I didn't think to test network performance (as suggested in another thread) until well into the process. Check networking first! QH ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] Check networking first?
Thanks for posting this! We see issues like this more often than you'd think. It's really important too because if you don't figure it out the natural inclination is to blame Ceph! :) Mark On 07/30/2015 12:50 PM, Quentin Hartman wrote: Just wanted to drop a note to the group that I had my cluster go sideways yesterday, and the root of the problem was networking again. Using iperf I discovered that one of my nodes was only moving data at 1.7Mb / s. Moving that node to a different switch port with a different cable has resolved the problem. It took awhile to track down because none of the server-side error metrics for disk or network showed anything was amiss, and I didn't think to test network performance (as suggested in another thread) until well into the process. Check networking first! QH ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] dropping old distros: el6, precise 12.04, debian wheezy?
I agree. For the existing stable series the distribution support should be continued. But for new releases (infernalis, jewel...) I see no problem dropping the older versions of the distributions. greetings Johannes Am 30.07.2015 um 16:39 schrieb Jon Meacham jomea...@adobe.com: If hammer and firefly bugfix releases will still be packaged for these distros, I don't see a problem with this. Anyone who is operating an existing LTS deployment on CentOS 6, etc. will continue to receive fixes for said LTS release. Jon From: ceph-users on behalf of Jan “Zviratko” Schermer Date: Thursday, July 30, 2015 at 8:29 AM To: Sage Weil Cc: ceph-devel, ceph-us...@ceph.com Subject: Re: [ceph-users] dropping old distros: el6, precise 12.04, debian wheezy? I understand your reasons, but dropping support for LTS release like this is not right. You should lege artis support every distribution the LTS release could have ever been installed on - that’s what the LTS label is for and what we rely on once we build a project on top of it CentOS 6 in particular is still very widely used and even installed, enterprise apps rely on it to this day. Someone out there is surely maintaining their LTS Ceph release on this distro and not having tested packages will hurt badly. We don’t want out project managers selecting EMC SAN over CEPH SDS because of such uncertainty, and you should benchmark yourself to those vendors, maybe... Every developer loves dropping support and concentrating on the bleeding edge interesting stuff but that’s not how it should work. Just my 2 cents... Jan On 30 Jul 2015, at 15:54, Sage Weil sw...@redhat.com wrote: As time marches on it becomes increasingly difficult to maintain proper builds and packages for older distros. For example, as we make the systemd transition, maintaining the kludgey sysvinit and udev support for centos6/rhel6 is a pain in the butt and eats up time and energy to maintain and test that we could be spending doing more useful work. Dropping them would mean: - Ongoing development on master (and future versions like infernalis and jewel) would not be tested on these distros. - We would stop building upstream release packages on ceph.com for new releases. - We would probably continue building hammer and firefly packages for future bugfix point releases. - The downstream distros would probably continue to package them, but the burden would be on them. For example, if Ubuntu wanted to ship Jewel on precise 12.04, they could, but they'd probably need to futz with the packaging and/or build environment to make it work. So... given that, I'd like to gauge user interest in these old distros. Specifically, CentOS6 / RHEL6 Ubuntu precise 12.04 Debian wheezy Would anyone miss them? In particular, dropping these three would mean we could drop sysvinit entirely and focus on systemd (and continue maintaining the existing upstart files for just a bit longer). That would be a relief. (The sysvinit files wouldn't go away in the source tree, but we wouldn't worry about packaging and testing them properly.) Thanks! sage ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] questions on editing crushmap for ceph cache tier
-BEGIN PGP SIGNED MESSAGE- Hash: SHA256 You are close... I've done it by creating a new SSD root in the CRUSH map, then put the SSD OSDs into a -ssd entry. I then created a new crush rule to choose from the SSD root, then have the tiering pool use that rule. If you look at the example in the document and think of ceph-osd-ssd-server-1 and ceph-osd-platter-server-1 as the same physical server with just logical separation, you can follow the rest of the example. You will need to modify either the ceph.conf to have the ssd OSDs have a different CRUSH map location or write a location script to do it automatically [1]. ceph.conf example: [osd.70] crush location = root=ssd host=nodez-ssd To do it programmaticly, you can use /usr/bin/ceph-crush-location as a base. I extended it by finding the device, then searching through hdparam to see if the rotation was greater than zero. If it wasn't then I output the hostname with the ssd portion, otherwise just the hostname. It was only a few lines of code, but I can't find it right now. I was waiting for ACLs to be modified so that I could query the data center location from our inventory management system (that was a few months ago) and I'm still waiting. [1] http://lists.ceph.com/pipermail/ceph-users-ceph.com/2015-January/045673.html -BEGIN PGP SIGNATURE- Version: Mailvelope v0.13.1 Comment: https://www.mailvelope.com wsFcBAEBCAAQBQJVunMXCRDmVDuy+mK58QAAq04P+QH95haU34fWZ5PsPIxv oY7JReywHwP3mWBO2XkaIg/l4AYV/HCBckBTSyp+GGFAtMeVEndiHCYTUf+F kQJjIk1jZoN+WTLnD8nsfDMrmVmforyGcG4Y399C4cCkBmeoU3jeGeKx+Unx dxiW6flRH5GPCazQdAAbXgb8InynUZ/EqmTY0FCDuLQ3CEXELuM8IReKwz0X 9LYqVIV+tdE1Ff2nDnLQlYYVpVv5K0y4TXBj8JzYH/41XbEws2GQnhb6b8zW aopUDC9RsNGtzWf4RDg8X3LDHrw7IBtAuJf+PHbcq3Y4cmPf5Z0TYiS1bqn1 19kj3EhDVVdW1KUue2S3GemyP0+bIypA/VDGzFXgv9g5oKN0bXPOjuFKAD2Q 7Rc2yoW70LACgL0a2KiRPRt8e6Jz5/vG6GijZvxTTZfkDPKHBOPPA3mFyAS4 FGmu39/q5VP7V+CepaKjbGNUWRzlzOOcc4ybk3dmktYEFOTw4QZLczBGJ8s4 I+AdYDjiQOAG3n3xixqRFOb4URjfKOrUbnHfNVQJU+qfYfV1RBLThhRjiv0v O+oiKiWuugZicHniTfHuOYePgxbs9eU2Hk8VRVk9ievXuRynDrH7D+IeUzUJ JGoj01YM60Ul1XJPWatMoM+435hcHrGd0rJ3bi91DOrZmT55X4jjdUA8z/3Y xMZE =Tqaw -END PGP SIGNATURE- Robert LeBlanc PGP Fingerprint 79A2 9CA4 6CC4 45DD A904 C70E E654 3BB2 FA62 B9F1 On Wed, Jul 29, 2015 at 9:21 PM, van chaofa...@owtware.com wrote: Hi, list, Ceph cache tier seems very promising for performance. According to http://ceph.com/docs/master/rados/operations/crush-map/#placing-different-pools-on-different-osds , I need to create a new pool based on SSD OSDs。 Currently, I’ve two servers with several HDD-based OSDs. I mean to add one SSD-based OSD for each server, and then use these two OSDs to build a cache pool. But I’ve found problems editing crushmap. The example in the link use two new hosts to build SSD OSDs and then create a new ruleset take the new hosts. But in my environment, I do not have new servers to use. Can I create a ruleset choose part of OSDs in a host? For example, as the crushmap shown below, osd.2 and osd.5 are new added SSD-based OSDs, how can I create a ruleset choose these two OSDs only and how can I avoid default ruleset to choose osd.2 and osd.5? Is this possible, or I have to add a new server to deploy cache tier? Thanks. host node0 { id -2 alg straw hash 0 item osd.0 weight 1.0 # HDD item osd.1 weight 1.0 # HDD item osd.2 weight 0.5 # SSD } host node1 { id -3 alg straw hash 0 item osd.3 weight 1.0 # HDD item osd.4 weight 1.0 # HDD item osd.5 weight 0.5 # SSD } root default { id -1 # do not change unnecessarily # weight 1.560 alg straw hash 0 # rjenkins1 item node0 weight 2.5 item node1 weight 2.5 } # typical ruleset rule replicated_ruleset { ruleset 0 type replicated min_size 1 max_size 10 step take default step chooseleaf firstn 0 type host step emit } van chaofa...@owtware.com ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] Recovery question
-BEGIN PGP SIGNED MESSAGE- Hash: SHA256 I'm glad you were able to recover. I'm sure you learned a lot about Ceph through the exercise (always seems to be the case for me with things). I'll look forward to your report so that we can include it in our operations manual, just in case. - Robert LeBlanc PGP Fingerprint 79A2 9CA4 6CC4 45DD A904 C70E E654 3BB2 FA62 B9F1 On Thu, Jul 30, 2015 at 12:41 PM, Peter Hinman wrote: For the record, I have been able to recover. Thank you very much for the guidance. I hate searching the web and finding only partial information on threads like this, so I'm going to document and post what I've learned as best I can in hopes that it will help someone else out in the future. -- Peter Hinman On 7/29/2015 5:15 PM, Robert LeBlanc wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA256 If you had multiple monitors, you should recover if possible more than 50% of them (they will need to form a quorum). If you can't, it is messy but, you can manually remove enough monitors to start a quorum. From /etc/ceph/ you will want the keyring and the ceph.conf at a minimim. The keys for the monitor I think are in the store.db which will let the monitors start, but the keyring has the admin key which lets you manage the cluster once you get it up. rbdmap is not needed for recovery (only automatically mounting RBDs at boot time), we can deal with that later if needed. - Robert LeBlanc PGP Fingerprint 79A2 9CA4 6CC4 45DD A904 C70E E654 3BB2 FA62 B9F1 On Wed, Jul 29, 2015 at 4:40 PM, Peter Hinman wrote: Ok - that is encouraging. I've believe I've got data from a previous monitor. I see files in a store.db dated yesterday, with a MANIFEST- file that is significantly greater than the MANIFEST-07 file listed for the current monitors. I've actually found data for two previous monitors. Any idea which one I should select? The one with the highest manifest number? The most recent time stamp? What files should I be looking for in /etc/conf? Just the keyring and rbdmap files? How important is it to use the same keyring file? -- Peter Hinman On 7/29/2015 3:47 PM, Robert LeBlanc wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA256 The default is /var/lib/ceph/mon/- (/var/lib/ceph/mon/ceph-mon1 for me). You will also need the information from /etc/ceph/ to reconstruct the data. I *think* you should be able to just copy this to a new box with the same name and IP address and start it up. I haven't actually done this, so there still may be some bumps. - Robert LeBlanc PGP Fingerprint 79A2 9CA4 6CC4 45DD A904 C70E E654 3BB2 FA62 B9F1 On Wed, Jul 29, 2015 at 3:44 PM, Peter Hinman wrote: Thanks Robert - Where would that monitor data (database) be found? -- Peter Hinman On 7/29/2015 3:39 PM, Robert LeBlanc wrote: -BEGIN PGP SIGNED MESSAGE- Hash: SHA256 If you built new monitors, this will not work. You would have to recover the monitor data (database) from at least one monitor and rebuild the monitor. The new monitors would not have any information about pools, OSDs, PGs, etc to allow an OSD to be rejoined. - Robert LeBlanc PGP Fingerprint 79A2 9CA4 6CC4 45DD A904 C70E E654 3BB2 FA62 B9F1 On Wed, Jul 29, 2015 at 2:46 PM, Peter Hinman wrote: Hi Greg - So at the moment, I seem to be trying to resolve a permission error. === osd.3 === Mounting xfs on stor-2:/var/lib/ceph/osd/ceph-3 2015-07-29 13:35:08.809536 7f0a0262e700 0 librados: osd.3 authentication error (1) Operation not permitted Error connecting to cluster: PermissionError failed: 'timeout 30 /usr/bin/ceph -c /etc/ceph/ceph.conf --name=osd.3 --keyring=/var/lib/ceph/osd/ceph-3/keyring osd crush create-or-move -- 3 3.64 host=stor-2 root=default' ceph-disk: Error: ceph osd start failed: Command '['/usr/sbin/service', 'ceph', '--cluster', 'ceph', 'start', 'osd.3']' returned non-zero exit status 1 ceph-disk: Error: One or more partitions failed to activate Is there a way to identify the cause of this PermissionError? I've copied the client.bootstrap-osd key from the output of ceph auth list, and pasted it into /var/lib/ceph/bootstrap-osd/ceph.keyring, but that has not resolve the error. But it sounds like you are saying that even once I get this resolved, I have no hope of recovering the data? -- Peter Hinman On 7/29/2015 1:57 PM, Gregory Farnum wrote: This sounds like you're trying to reconstruct a cluster after destroying the monitors. That is...not going to work well. The monitors define the cluster and you can't move OSDs into different clusters. We have ideas for how to reconstruct monitors and it can be done manually with a lot of hassle, but the process isn't written down and there aren't really fools I help with it. :/ -Greg On Wed, Jul 29, 2015 at 5:48 PM Peter
Re: [ceph-users] Elastic-sized RBD planned?
-BEGIN PGP SIGNED MESSAGE- Hash: SHA256 I'll take a stab at this. I don't think it will be a feature that you will find in Ceph due to the fact that Ceph doesn't really understand what is going on inside the RBD. There are too many technologies that can use RBD that it is not feasible to try and support something like this. You can however have a service that runs in your VM which monitors free space. Then the free space gets too low, it can call to a service you write which will then expand the RBD on the fly and then the VM itself and resize the partitions and the file system after the RBD is expanded. I'm not sure how CephFS fits into all of this as it is different than RBD. CephFS is to NFS as RBD is to SAN block storage. If I misunderstood the question, please clarify. -BEGIN PGP SIGNATURE- Version: Mailvelope v0.13.1 Comment: https://www.mailvelope.com wsFcBAEBCAAQBQJVunUCCRDmVDuy+mK58QAAwGcP/0Y5gz4flZ0XuaXdKlez iogX9QsAoPk9s8fd0vpc3Prlhx/YODgkqqm35iJh5PaDEqb6njZwe3CR0WBh mba6xAjSGe9D8dzkLP5cCvRSlkVexddAfj5K/M+JkjWyQhlq4TQcu3CSBo8Q 1pyI6dDwWNl8ScCu4/PN4Bl2OD9favEs8tXjNJ5/mhZWFjSN7t8/LLqsUNu6 AlMp2hFFfb1f0ky7EI/Hg3uw0+BbGDn/N0oxDyZqH7Wqnp/5L4kRAkZkXsGh T7a4KxReNu/5eg2Tef83h7AAeRTSDTEw+38ToGnOIGpXYCxKdWF878Xs7tVa +jPUds7aSNazaRB9nSPYiIxXBcTRKN2VFqhFNyQ/6CrEvsjkZQKkQfjJgNtL eg0hmjS1X0QryapX7xhfz+Apx369Pkitm8UosyPIwEPnMuVqwVN5VwDTDkub FlGNHX+b1/NDgZDpWF+b5gOErHMW8kWRNt/+2i5pXj0ZjDADmrQn+Hd9G0Hx g1dot64vLogcvcyt0C+fLicF9xlddU/Zuz7VZLyIOH1KSVhABK1RaI8+Zws6 ZWriDFal0ztd0BNEQlCqtlo4hyY/AVies9qB6V4sUL0UuEL/+iTj71/VNh09 RJIURK6KySwxtW97pMGGafw5xPOjpxnm75D6AmovZ6WV68GSxlNpTY/V7CPH dW/U =h5n5 -END PGP SIGNATURE- Robert LeBlanc PGP Fingerprint 79A2 9CA4 6CC4 45DD A904 C70E E654 3BB2 FA62 B9F1 On Wed, Jul 29, 2015 at 11:32 PM, Shneur Zalman Mattern shz...@eimsys.co.il wrote: Hi to all! Perhaps, somebody already thought about, but my Googling had no results. How can I do RBD that will grow on demand of VM/client disk space. Are there in Ceph some options for this? Is it planned to do? Is it utopic idea? Is this client need CephFS already? Thanks, Shneur This footnote confirms that this email message has been scanned by PineApp Mail-SeCure for the presence of malicious code, vandals computer viruses. ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] Crash and question
Hi, I tried to add a new monitor, but now I was unable to use ceph command after doing ceph-deploy mon create myhostname I’ve got : # ceph status 2015-07-30 10:42:39.682038 7f7b16d90700 0 librados: client.admin authentication error (1) Operation not permitted Error connecting to cluster: PermissionError Could you help me to fix and how to change keys please ? Thank a lot in advance, and sorry I’m a newbie on this topic. K Le 30 juil. 2015 à 11:39, Khalid Ahsein kahs...@gmail.com a écrit : Good morning christian, thank you for your quick response. so I need to upgrade to 64 GB or 96 GB to be more secure ? And sorry I though that 2 monitors was the minimum. We will work to add a new host quickly. About osd_pool_default_min_size should I change something for the future ? thank you again K Le 30 juil. 2015 à 11:12, Christian Balzer ch...@gol.com mailto:ch...@gol.com a écrit : Hello, On Thu, 30 Jul 2015 10:55:30 +0200 Khalid Ahsein wrote: Hello everybody, I’m running since 4 months a ceph cluster configured with two monitors : 1 host : 16GB RAM - 12x 4TB disks - 12 OSD - 1 monitor - RAID-1 for system 1 host : 16GB RAM - 12x 4TB disks - 12 OSD - 1 monitor - RAID-1 for system Too little RAM, just 2 monitors, just 2 nodes... This night I’ve encountered an issue with the crash of the first host. My first question is why with 1 host down, all my cluster was down (unable to do ceph status — hang command) and all my rbd was stuck without possibility to R/W. Re-read the documentation, you need at least 3 monitors to survive the loss of one (monitor) node. Your osd_pool_default_min_size would have left in a usable situation, 2 nodes is really a minimal case. I rebooted the first host, and 2 hours later the second go down with the same issue (all rbd down and ceph hang). After reboot, here is ceph status : # ceph status cluster 9c29f469-7bad-4b64-97bf-3fbb1bbc0c5f health HEALTH_ERR 3 pgs inconsistent 1 pgs peering 1 pgs stuck inactive 1 pgs stuck unclean 36 requests are blocked 32 sec 928 scrub errors clock skew detected on mon.drt-becks monmap e1: 2 mons at {drt-becks=172.16.21.6:6789/0,drt-marco=172.16.21.4:6789/0} election epoch 26, quorum 0,1 drt-marco,drt-becks osdmap e961: 24 osds: 24 up, 24 in pgmap v2532968: 400 pgs, 1 pools, 512 GB data, 130 kobjects 1039 GB used, 88092 GB / 89177 GB avail 393 active+clean 3 active+clean+scrubbing+deep 3 active+clean+inconsistent 1 peering client io 57290 B/s wr, 7 op/s You will want to: a) fix your NTP, clock skew. b) check your logs about the scrub errors c) same for the stuck requests Also I found this error on DMESG about the crash : Message from syslogd@drt-marco at Jul 30 04:03:57 ... kernel:[4876519.657178] BUG: soft lockup - CPU#7 stuck for 22s! [btrfs-cleaner:32713] All my volumes are on BTRFS, maybe it was not a good idea ? Depending on your OS, kernel version, most definitely. Plenty of BTRFS problems in the ML archives to be found. Christian -- Christian BalzerNetwork/Systems Engineer ch...@gol.com mailto:ch...@gol.com Global OnLine Japan/Fusion Communications http://www.gol.com/ http://www.gol.com/ ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] Weird behaviour of cephfs with samba
Gregory Farnum greg@... writes: You can mount subtrees with the -r option to ceph-fuse. Yay! That did the trick to properly mount via fuse. And I can confirm that the directory list results are now stable both locally and via samba. Once you've started it up you should find a file like client.admin.[0-9]*.asok in (I think?) /var/run/ceph. You can run ceph --admin-daemon /var/run/ceph/{client_asok} status and provide the output to see if it's doing anything useful. Or set debug client = 20 in the config and then upload the client log file either publicly or with ceph-post-file and I'll take a quick look to see what's going on. root@gru:/mnt# ceph --admin-daemon /var/run/ceph/ceph-client.admin.asok status { metadata: { ceph_sha1: 5fb85614ca8f354284c713a2f9c610860720bbf3, ceph_version: ceph version 0.94.2 (5fb85614ca8f354284c713a2f9c610860720bbf3), entity_id: admin, hostname: gru, mount_point: \/mnt\/test }, dentry_count: 0, dentry_pinned_count: 0, inode_count: 1, mds_epoch: 1075, osd_epoch: 2391, osd_epoch_barrier: 0 } It seems like if cephfs was only ever mounted with a non-root basepath, the cephfs root does not contain a directory entry for that base path. At least that's what I guess from dentry_count: 0. Mmm, that looks like a Samba config issue which unfortunately I don't know much about. Perhaps you need to install these modules individually? It looks like our nightly tests are just getting the Ceph VFS installed by default. :/ Ubuntu's samba package isn't configured with --with-libcephfs and therefore simply doesn't ship ceph.so. I am currently trying to recompile it with that flag. If that doesn't work, that's no biggie, because the re-export of the ceph-fuse mounted directory behaves flawlessly now. Thanks a ton for your help! Joerg Henne ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] which kernel version can help avoid kernel client deadlock
Date: Thu, 30 Jul 2015 11:37:37 +0300 Subject: Re: [ceph-users] which kernel version can help avoid kernel client deadlock From: idryo...@gmail.com To: zhangz.da...@outlook.com CC: chaofa...@owtware.com; ceph-users@lists.ceph.com On Thu, Jul 30, 2015 at 10:29 AM, Z Zhang zhangz.da...@outlook.com wrote: Subject: Re: [ceph-users] which kernel version can help avoid kernel client deadlock From: chaofa...@owtware.com Date: Thu, 30 Jul 2015 13:16:16 +0800 CC: idryo...@gmail.com; ceph-users@lists.ceph.com To: zhangz.da...@outlook.com On Jul 30, 2015, at 12:48 PM, Z Zhang zhangz.da...@outlook.com wrote: We also hit the similar issue from time to time on centos with 3.10.x kernel. By iostat, we can see kernel rbd client's util is 100%, but no r/w io, and we can't umount/unmap this rbd client. After restarting OSDs, it will become normal. 3.10.x is rather vague, what is the exact version you saw this on? Can you provide syslog logs (I'm interested in dmesg)? The kernel version should be 3.10.0. I don't have sys logs at hand. It is not easily reproduced, and it happened at very low memory situation. We are running DB instances over rbd as storage. DB instances will use lot of memory when running high concurrent rw, and after running for a long time, rbd might hit this problem, but not always. Enabling rbd log makes our system behave strange during our test. I back-ported one of your fixes: https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/drivers/block/rbd.c?id=5a60e87603c4c533492c515b7f62578189b03c9c So far test looks fine for few days, but still under observation. So want to know if there are some other fixes? Thanks, Ilya ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] which kernel version can help avoid kernel client deadlock
On Thu, Jul 30, 2015 at 12:46 PM, Z Zhang zhangz.da...@outlook.com wrote: Date: Thu, 30 Jul 2015 11:37:37 +0300 Subject: Re: [ceph-users] which kernel version can help avoid kernel client deadlock From: idryo...@gmail.com To: zhangz.da...@outlook.com CC: chaofa...@owtware.com; ceph-users@lists.ceph.com On Thu, Jul 30, 2015 at 10:29 AM, Z Zhang zhangz.da...@outlook.com wrote: Subject: Re: [ceph-users] which kernel version can help avoid kernel client deadlock From: chaofa...@owtware.com Date: Thu, 30 Jul 2015 13:16:16 +0800 CC: idryo...@gmail.com; ceph-users@lists.ceph.com To: zhangz.da...@outlook.com On Jul 30, 2015, at 12:48 PM, Z Zhang zhangz.da...@outlook.com wrote: We also hit the similar issue from time to time on centos with 3.10.x kernel. By iostat, we can see kernel rbd client's util is 100%, but no r/w io, and we can't umount/unmap this rbd client. After restarting OSDs, it will become normal. 3.10.x is rather vague, what is the exact version you saw this on? Can you provide syslog logs (I'm interested in dmesg)? The kernel version should be 3.10.0. I don't have sys logs at hand. It is not easily reproduced, and it happened at very low memory situation. We are running DB instances over rbd as storage. DB instances will use lot of memory when running high concurrent rw, and after running for a long time, rbd might hit this problem, but not always. Enabling rbd log makes our system behave strange during our test. I back-ported one of your fixes: https://git.kernel.org/cgit/linux/kernel/git/torvalds/linux.git/commit/drivers/block/rbd.c?id=5a60e87603c4c533492c515b7f62578189b03c9c So far test looks fine for few days, but still under observation. So want to know if there are some other fixes? I'd suggest following 3.10 stable series (currently at 3.10.84). The fix you backported is crucial in low memory situations, so I wouldn't be surprised if it alone fixed your problem. (It is not in 3.10.84, I assume it'll show up in 3.10.85 - for now just apply your backport.) Thanks, Ilya ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] Squeeze packages for 0.94.2
Hello, On Thu, 30 Jul 2015 08:49:16 + Sebastian Köhler wrote: Hello, it seems that there are no Debian Squeeze packages in the repository for the current Hammer version. Is this an oversight or is there another reason those are not provided? Most likely because it's 2 versions behind the curve? Doing a fresh install on Squeeze is probably not something people consider to be likely or wise. Is there any reason you can't use Wheezy or Jessie? I can think of MANY reasons for not going to Jessie, main point for me is the lack of pacemaker, however Wheezy would make a good platform for the time being, seemingly also with LTS support. Christian Sebastian ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com -- Christian BalzerNetwork/Systems Engineer ch...@gol.com Global OnLine Japan/Fusion Communications http://www.gol.com/ ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] How to identify MDS client failing to respond to capability release?
Hi John, thanks a lot - I was indeed able to identify the machine in question. As for the kernel, we'll certainly update to a newer kernel (3.16 and later 3.19) for the Ubuntu 14.04 clients. For the 12.04 clients, we'll have to see, but these machines will be phased out over time anyhow. I'd like to avoid the FUSE client, as the performance is so much worse than with the kernel client. Thanks again! Oliver On 30.07.2015 09:50, John Spray wrote: For sufficiently recent clients we do this for you (clients send some metadata like hostname, which is used in the MDS to generate an easier-to-understand identifier). To do it by hand, use the admin socket command ceph daemon mds.id session ls command, and look out for the client IP addresses in the address part of each entry. By the way, there have been various other bugs in 3.13-ish kernels, so you should consider using either a more recent kernel or a fuse client. John On 30/07/15 08:32, Oliver Schulz wrote: Hello Ceph Experts, lately, ceph status on our cluster often states: mds0: Client CLIENT_ID failing to respond to capability release How can I identify which client is at fault (hostname or IP address) from the CLIENT_ID? What could be the source of the failing to respond to capability release - Linux kernel on the client too old? We use ceph-0.94.2 on the cluster, and the CephFS kernel client on the clients (kernel 3.13.0 on Ubuntu 12.04 and Ubuntu 14.04). But it's possible that there's a machine with an older Kernel around somewhere ... Cheers and thanks, Oliver ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] A cache tier issue with rate only at 20MB/s when data move from cold pool to hot pool
On 06/16/2015 01:17 PM, Kenneth Waegeman wrote: Hi! We also see this at our site: When we cat a large file from cephfs to /dev/null, we get about 10MB/s data transfer. I also do not see a system resource bottleneck. Our cluster consists of 14 servers with each 16 disks, together forming a EC coded pool. We also have 2SSDs per server for the cache. Running 0.94.1 Hi, Does someone has an idea about this? Is there some debugging or testing we can do to find the problem here? Thank you! Kenneth So we are having the same question. Our cache pool is in writeback mode, can it help to set it in readonly for this? Kenneth On 06/16/2015 12:58 PM, liukai wrote: Hi all, A cache tier, 2 hot node with 8 ssd osd, and 2 cold node with 24 sata osd. The public network rate is 1Mb/s and cluster network rate is 1000Mb/s. Using fuse-client to access the files. The issue is: When the files are in hot pool, the copy rate is very fash. But when the files are only in cold pool, the rate only reach 20MB/s. I known that when files are not in the hot pool, the files should been copy from cold pool to hot pool first, and then from hot pool to client. But the cpu, ram and network seems not the bottleneck. and will be the cause of system design? Are there some params to adjust to improve the rate from cold pool to hot pool? Thanks 2015-06-16 liukai ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] Crash and question
Good morning christian, thank you for your quick response. so I need to upgrade to 64 GB or 96 GB to be more secure ? And sorry I though that 2 monitors was the minimum. We will work to add a new host quickly. About osd_pool_default_min_size should I change something for the future ? thank you again K Le 30 juil. 2015 à 11:12, Christian Balzer ch...@gol.com a écrit : Hello, On Thu, 30 Jul 2015 10:55:30 +0200 Khalid Ahsein wrote: Hello everybody, I’m running since 4 months a ceph cluster configured with two monitors : 1 host : 16GB RAM - 12x 4TB disks - 12 OSD - 1 monitor - RAID-1 for system 1 host : 16GB RAM - 12x 4TB disks - 12 OSD - 1 monitor - RAID-1 for system Too little RAM, just 2 monitors, just 2 nodes... This night I’ve encountered an issue with the crash of the first host. My first question is why with 1 host down, all my cluster was down (unable to do ceph status — hang command) and all my rbd was stuck without possibility to R/W. Re-read the documentation, you need at least 3 monitors to survive the loss of one (monitor) node. Your osd_pool_default_min_size would have left in a usable situation, 2 nodes is really a minimal case. I rebooted the first host, and 2 hours later the second go down with the same issue (all rbd down and ceph hang). After reboot, here is ceph status : # ceph status cluster 9c29f469-7bad-4b64-97bf-3fbb1bbc0c5f health HEALTH_ERR 3 pgs inconsistent 1 pgs peering 1 pgs stuck inactive 1 pgs stuck unclean 36 requests are blocked 32 sec 928 scrub errors clock skew detected on mon.drt-becks monmap e1: 2 mons at {drt-becks=172.16.21.6:6789/0,drt-marco=172.16.21.4:6789/0} election epoch 26, quorum 0,1 drt-marco,drt-becks osdmap e961: 24 osds: 24 up, 24 in pgmap v2532968: 400 pgs, 1 pools, 512 GB data, 130 kobjects 1039 GB used, 88092 GB / 89177 GB avail 393 active+clean 3 active+clean+scrubbing+deep 3 active+clean+inconsistent 1 peering client io 57290 B/s wr, 7 op/s You will want to: a) fix your NTP, clock skew. b) check your logs about the scrub errors c) same for the stuck requests Also I found this error on DMESG about the crash : Message from syslogd@drt-marco at Jul 30 04:03:57 ... kernel:[4876519.657178] BUG: soft lockup - CPU#7 stuck for 22s! [btrfs-cleaner:32713] All my volumes are on BTRFS, maybe it was not a good idea ? Depending on your OS, kernel version, most definitely. Plenty of BTRFS problems in the ML archives to be found. Christian -- Christian BalzerNetwork/Systems Engineer ch...@gol.com mailto:ch...@gol.com Global OnLine Japan/Fusion Communications http://www.gol.com/ http://www.gol.com/ ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
[ceph-users] Crash and question
Hello everybody, I’m running since 4 months a ceph cluster configured with two monitors : 1 host : 16GB RAM - 12x 4TB disks - 12 OSD - 1 monitor - RAID-1 for system 1 host : 16GB RAM - 12x 4TB disks - 12 OSD - 1 monitor - RAID-1 for system This night I’ve encountered an issue with the crash of the first host. My first question is why with 1 host down, all my cluster was down (unable to do ceph status — hang command) and all my rbd was stuck without possibility to R/W. I rebooted the first host, and 2 hours later the second go down with the same issue (all rbd down and ceph hang). After reboot, here is ceph status : # ceph status cluster 9c29f469-7bad-4b64-97bf-3fbb1bbc0c5f health HEALTH_ERR 3 pgs inconsistent 1 pgs peering 1 pgs stuck inactive 1 pgs stuck unclean 36 requests are blocked 32 sec 928 scrub errors clock skew detected on mon.drt-becks monmap e1: 2 mons at {drt-becks=172.16.21.6:6789/0,drt-marco=172.16.21.4:6789/0} election epoch 26, quorum 0,1 drt-marco,drt-becks osdmap e961: 24 osds: 24 up, 24 in pgmap v2532968: 400 pgs, 1 pools, 512 GB data, 130 kobjects 1039 GB used, 88092 GB / 89177 GB avail 393 active+clean 3 active+clean+scrubbing+deep 3 active+clean+inconsistent 1 peering client io 57290 B/s wr, 7 op/s Also I found this error on DMESG about the crash : Message from syslogd@drt-marco at Jul 30 04:03:57 ... kernel:[4876519.657178] BUG: soft lockup - CPU#7 stuck for 22s! [btrfs-cleaner:32713] All my volumes are on BTRFS, maybe it was not a good idea ? Thanks a lot for your help, on the bottom more hardware information K # cat /proc/cpuinfo processor : 0 vendor_id : GenuineIntel cpu family : 6 model : 26 model name : Intel(R) Xeon(R) CPU E5506 @ 2.13GHz stepping: 5 microcode : 0x19 cpu MHz : 2133.433 cache size : 4096 KB physical id : 1 siblings: 4 core id : 0 cpu cores : 4 apicid : 16 initial apicid : 16 fpu : yes fpu_exception : yes cpuid level : 11 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx rdtscp lm constant_tsc arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc aperfmperf pni dtes64 monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr pdcm dca sse4_1 sse4_2 popcnt lahf_lm dtherm tpr_shadow vnmi flexpriority ept vpid bogomips: 4266.86 clflush size: 64 cache_alignment : 64 address sizes : 40 bits physical, 48 bits virtual power management: processor : 1 vendor_id : GenuineIntel cpu family : 6 model : 26 model name : Intel(R) Xeon(R) CPU E5506 @ 2.13GHz stepping: 5 microcode : 0x19 cpu MHz : 2133.433 cache size : 4096 KB physical id : 0 siblings: 4 core id : 0 cpu cores : 4 apicid : 0 initial apicid : 0 fpu : yes fpu_exception : yes cpuid level : 11 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx rdtscp lm constant_tsc arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc aperfmperf pni dtes64 monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr pdcm dca sse4_1 sse4_2 popcnt lahf_lm dtherm tpr_shadow vnmi flexpriority ept vpid bogomips: 4266.74 clflush size: 64 cache_alignment : 64 address sizes : 40 bits physical, 48 bits virtual power management: processor : 2 vendor_id : GenuineIntel cpu family : 6 model : 26 model name : Intel(R) Xeon(R) CPU E5506 @ 2.13GHz stepping: 5 microcode : 0x19 cpu MHz : 2133.433 cache size : 4096 KB physical id : 1 siblings: 4 core id : 1 cpu cores : 4 apicid : 18 initial apicid : 18 fpu : yes fpu_exception : yes cpuid level : 11 wp : yes flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx rdtscp lm constant_tsc arch_perfmon pebs bts rep_good nopl xtopology nonstop_tsc aperfmperf pni dtes64 monitor ds_cpl vmx est tm2 ssse3 cx16 xtpr pdcm dca sse4_1 sse4_2 popcnt lahf_lm dtherm tpr_shadow vnmi flexpriority ept vpid bogomips: 4266.86 clflush size: 64 cache_alignment : 64 address sizes : 40 bits physical, 48 bits virtual power management: processor : 3 vendor_id : GenuineIntel cpu family : 6 model : 26 model name : Intel(R) Xeon(R) CPU E5506 @ 2.13GHz stepping: 5 microcode : 0x19 cpu MHz : 2133.433 cache
Re: [ceph-users] Crash and question
Hello, On Thu, 30 Jul 2015 10:55:30 +0200 Khalid Ahsein wrote: Hello everybody, I’m running since 4 months a ceph cluster configured with two monitors : 1 host : 16GB RAM - 12x 4TB disks - 12 OSD - 1 monitor - RAID-1 for system 1 host : 16GB RAM - 12x 4TB disks - 12 OSD - 1 monitor - RAID-1 for system Too little RAM, just 2 monitors, just 2 nodes... This night I’ve encountered an issue with the crash of the first host. My first question is why with 1 host down, all my cluster was down (unable to do ceph status — hang command) and all my rbd was stuck without possibility to R/W. Re-read the documentation, you need at least 3 monitors to survive the loss of one (monitor) node. Your osd_pool_default_min_size would have left in a usable situation, 2 nodes is really a minimal case. I rebooted the first host, and 2 hours later the second go down with the same issue (all rbd down and ceph hang). After reboot, here is ceph status : # ceph status cluster 9c29f469-7bad-4b64-97bf-3fbb1bbc0c5f health HEALTH_ERR 3 pgs inconsistent 1 pgs peering 1 pgs stuck inactive 1 pgs stuck unclean 36 requests are blocked 32 sec 928 scrub errors clock skew detected on mon.drt-becks monmap e1: 2 mons at {drt-becks=172.16.21.6:6789/0,drt-marco=172.16.21.4:6789/0} election epoch 26, quorum 0,1 drt-marco,drt-becks osdmap e961: 24 osds: 24 up, 24 in pgmap v2532968: 400 pgs, 1 pools, 512 GB data, 130 kobjects 1039 GB used, 88092 GB / 89177 GB avail 393 active+clean 3 active+clean+scrubbing+deep 3 active+clean+inconsistent 1 peering client io 57290 B/s wr, 7 op/s You will want to: a) fix your NTP, clock skew. b) check your logs about the scrub errors c) same for the stuck requests Also I found this error on DMESG about the crash : Message from syslogd@drt-marco at Jul 30 04:03:57 ... kernel:[4876519.657178] BUG: soft lockup - CPU#7 stuck for 22s! [btrfs-cleaner:32713] All my volumes are on BTRFS, maybe it was not a good idea ? Depending on your OS, kernel version, most definitely. Plenty of BTRFS problems in the ML archives to be found. Christian -- Christian BalzerNetwork/Systems Engineer ch...@gol.com Global OnLine Japan/Fusion Communications http://www.gol.com/ ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] Squeeze packages for 0.94.2
July 30 2015 11:05 AM, Christian Balzer ch...@gol.com wrote: Is there any reason you can't use Wheezy or Jessie? Our cluster is running on trusty however nearly all our clients are running on squeeze and can not be updated for compatibility reasons in the short term. Packages of older Hammer versions were provided so we assumed future releases of Hammer would be provided for Debian 6. Sebastian ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com