Re: [ceph-users] Number of ODS per host
Hi, I think it depends of your total number of OSDs. I take my case : I have 8 OSD per host, and 5 host. If one host crash, I loose 20% of the cluster, and have a huge amount of data to rebalance. For fault tolerance, it was not a good idea. Le mardi 05 mars 2013 à 02:39 -0800, waed Albataineh a écrit : Hi there, i believe for the quick start of Ceph we will get two OSDs, even it's not recommended i wanna increase them. My question is it gonna ended bad if i end up with 10 OSDs per host ?? and for these increment i must manipulate the configuration file, right ?? finally if i finished the installation then i realize i need to change sth on the configuration file is it possible or it will crash ?? ___ 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] A problem about RBD moudule
I installed ceph 0.56 on Red Hat Enterprise Linux Server release 6.0 (Santiago). It works well. but, when I follow the block device quick start, I get a problem. [root@crt24 ceph]# rbd create foo --size 4096 [root@crt24 ceph]# rados ls -p rbd foo.rbd rbd_directory [root@crt24 ceph]# modeprobe rbd -- Then. An error occurred! -- [root@crt24 ceph]# rbd map foo --pool rbd --name client.admin rbd: /sys/bus/rbd/add does not exist! Did you run 'modprobe rbd' or is your rbd module too old? rbd: add failed: (2) No such file or directory -- So I check lsmod first. but no rbd moudule found. -- [root@crt24 fs]# lsmod |grep rbd [root@crt24 fs]# -- Then, I try to find rbd.ko. but no rbd.ko too. -- [root@crt24 ceph]# cd /lib/modules/2.6.34.14/kernel/drivers/block [root@crt24 fs]# ls aoe cciss.ko cryptoloop.ko drbd nbd.ko sx8.ko umem.ko brd.ko cpqarray.ko DAC960.ko floppy.ko pktcdvd.ko ub.ko So my idea is that recompile the kernel. And compile the RBD module into the kernel. While, unfortunately,I can't find the RBD module when I “make menuconfig”. Now, I want to know that Is there another solution to add the RBD module, or what I exactly went wrong! thanks a lot! ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] I/O Speed Comparisons
On 03/06/2013 07:28 AM, Wolfgang Hennerbichler wrote: Hi, I've set up (latest bobtail) ceph between 2 Machines, 2 OSDs each. Then I've started up a VM backed by LVM with windows something in it, and i ran the crappy ATTO Disk Benchmark to get performance information (my customer wants me to do it like that, sorry). So with 4MB Blocks being written and read, this tool claims that this certain setup writes with 6.5 MB/s and reads with 182MB/s. Done the same with ceph, same virtual machine, same everything (no other stuff is running in this testing environment), except for this in libvirt: disk type='network' device='disk' driver name='qemu' type='raw'/ source protocol='rbd' name='rd/korfu:rbd_cache=1' host name='rd-clusternode21' port='6789'/ host name='rd-clusternode22' port='6789'/ /source target dev='hda' bus='ide'/ address type='drive' controller='0' bus='0' unit='0'/ /disk Write-Speed is way better than LVM (probably because of caching): 111MB/s. But Read-Speed is really bad: 37MB/s (vs. 182MB/s with lvm). The virtual Machine can somehow no longer maintain time, it drifts by various minutes. Ceph setup is pretty basic. Can anybody give me some hints where I could turn some knobs? I'd rather trade some write-speed to get better read-speed. If you are doing sequential reads, you may benefit by increasing the read_ahead_kb value for each device in /sys/block/device/queue on the OSD hosts. Wolfgang ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] A problem about RBD moudule
Hi, it would be much easier to grab a prepackaged mainline kernel from elrepo, which includes the Ceph-related modules. The distro packaged kernel in RHEL / Cent / Sci doesn't provide any of these. http://elrepo.org/linux/kernel/el6/x86_64/RPMS/ On Wed, Mar 6, 2013 at 3:05 AM, Jiang Long longjiang2...@gmail.com wrote: I installed ceph 0.56 on Red Hat Enterprise Linux Server release 6.0 (Santiago). It works well. but, when I follow the block device quick start, I get a problem. [root@crt24 ceph]# rbd create foo --size 4096 [root@crt24 ceph]# rados ls -p rbd foo.rbd rbd_directory [root@crt24 ceph]# modeprobe rbd -- Then. An error occurred! -- [root@crt24 ceph]# rbd map foo --pool rbd --name client.admin rbd: /sys/bus/rbd/add does not exist! Did you run 'modprobe rbd' or is your rbd module too old? rbd: add failed: (2) No such file or directory -- So I check lsmod first. but no rbd moudule found. -- [root@crt24 fs]# lsmod |grep rbd [root@crt24 fs]# -- Then, I try to find rbd.ko. but no rbd.ko too. -- [root@crt24 ceph]# cd /lib/modules/2.6.34.14/kernel/drivers/block [root@crt24 fs]# ls aoe cciss.ko cryptoloop.ko drbd nbd.ko sx8.ko umem.ko brd.ko cpqarray.ko DAC960.ko floppy.ko pktcdvd.ko ub.ko So my idea is that recompile the kernel. And compile the RBD module into the kernel. While, unfortunately,I can't find the RBD module when I “make menuconfig”. Now, I want to know that Is there another solution to add the RBD module, or what I exactly went wrong! thanks a lot! ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com -- __ Dino Yancey 2GNT.com Admin ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
Re: [ceph-users] bucket/user stats for billing
On Wed, Mar 6, 2013 at 2:25 PM, Gandalf Corvotempesta gandalf.corvotempe...@gmail.com wrote: 2013/3/6 Gandalf Corvotempesta gandalf.corvotempe...@gmail.com: Seems perfect. Thank you. But there is no way to get users list... You can do: rados -p .users.uid ls | grep -v .buckets Not very clean, and not too efficient. I'm working on some api that'll make it cleaner, but it's not going to be ready any time soon. Yehuda ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
[ceph-users] Wildcard for rgw keystone accepted roles
Is there a way to allow all roles (aside from listing them all) access to radosgw when using keystone? If new roles get added, I want them to automatically have access. ___ ceph-users mailing list ceph-users@lists.ceph.com http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com