It totally depends upon why kind of hardware you are using in your ceph 
cluster.  you should always choose the right hardware for your cluster keeping 
your performance needs in mind.

Can you tell us about the hardware you have or planning to purchase , also what 
kind of workload your setup would have.

- karan Singh -


On 24 Apr 2014, at 11:34, *sm1Ly <st.uz...@gmail.com> wrote:

> Hi Karan
> 
> thx for u cooperation.
> I got some help on #ceph, so now I got mon[123] and mds[123] and ods[12345].
> also I got one client on centos6,5 with elrepo lt kernel, which provides me 
> nonfuse mount of my cephfs.
> 
> I got some simple cases for test. like bonnie++ and this:
> for i in `seq 0 256`; do time mkdir $i ; done
> 
> cd 256
> 
> for i in `seq 0 256`; do time mkdir $i ; done
> 
> for i in `seq 0 5`; do time ls > /dev/null ; done
> 
> time find . > /dev/null
> 
> so I got next results:
> 
> local fs
> real    0m0.007s
> user    0m0.002s
> sys     0m0.004s
> 
> ceph fs
> real    0m0.394s
> user    0m0.008s
> sys     0m0.026s
> 
> and bonnie:
> 
> local
> 
> Version  1.96       ------Sequential Output------ --Sequential Input- 
> --Random-
> Concurrency   8     -Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- 
> --Seeks--
> Machine        Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP  /sec 
> %CP
> web11.abboom.wor 4G   585  99 56297   9 45337   7  3009  99 112865   7  1472  
> 31
> Latency             22491us     451ms     454ms   12023us     110ms   29311us
> Version  1.96       ------Sequential Create------ --------Random 
> Create--------
> web11.abboom.world  -Create-- --Read--- -Delete-- -Create-- --Read--- 
> -Delete--
>               files  /sec %CP  /sec %CP  /sec %CP  /sec %CP  /sec %CP  /sec 
> %CP
>                 100 47361  66 +++++ +++ 63536  68 52393  72 +++++ +++ 61602  
> 68
> Latency               911us     667us    2674us     240us     181us    1976us
> 1.96,1.96,web11.abboom.world,8,1398170891,4G,,585,99,56297,9,45337,7,3009,99,112865,7,1472,31,100,,,,,47361,66,+++++,+++,63536,68,52393,72,+++++,+++,61602,68,22491us,451ms,454ms,12023us,110ms,29311us,911us,667us,2674us,240us,181us,1976us
> 
> 
> ceph
> 
> Version  1.96       ------Sequential Output------ --Sequential Input- 
> --Random-
> Concurrency   8     -Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- 
> --Seeks--
> Machine        Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP  /sec 
> %CP
> web11.abboom.wor 4G   489  97 15635   2 15565   2   839  99 319882  21 817.0  
> 16
> Latency               124ms   35589us    9818ms   15806us     730ms   48226us
> Version  1.96       ------Sequential Create------ --------Random 
> Create--------
> web11.abboom.world  -Create-- --Read--- -Delete-- -Create-- --Read--- 
> -Delete--
>               files  /sec %CP  /sec %CP  /sec %CP  /sec %CP  /sec %CP  /sec 
> %CP
>                 100   118   0 52002  11    66   0    97   0  8582   9    60   > 0
> Latency             11718ms   32137us    8566ms    6235ms     204ms   24989ms
> 1.96,1.96,web11.abboom.world,8,1398170818,4G,,489,97,15635,2,15565,2,839,99,319882,21,817.0,16,100,,,,,118,0,52002,11,66,0,97,0,8582,9,60,0,124ms,35589us,9818ms,15806us,730ms,48226us,11718ms,32137us,8566ms,6235ms,204ms,24989ms
> 
> 
> so I got some questions now.
> all me scheme works on hp storage -> iscsi -> vmware vmfs -> vitrual machine 
> with ods or mon or mds.
> 
> if I will use just hardware. I mean what if I got some hw with linux install 
> on it and 20hdds where ods = hdd will it be faster?
> 
> or I need some advice how to tune cephfs for faster works.
> 
> 
> 
> 
> 
> 2014-04-23 11:48 GMT+04:00 Karan Singh <karan.si...@csc.fi>:
> Hi Alexander
> 
> Try adding  your monitor details in  /etc/ceph/ceph.conf file (please check 
> for typos)
> 
> 
> [mon]
> 
> [mon.nfs2.abboom.world]
>     host = nfs2.abboom.world
>     mon addr = 10.60.0.111:6789
> 
> [mon.nfs3.abboom.world]
>     host = nfs3.abboom.world
>     mon addr = 10.60.0.112:6789
> 
> [mon.nfs4.abboom.world]
>     host = nfs4.abboom.world
>     mon addr = 10.60.0.113:6789
> 
> 
> - karan -
> 
> On 21 Apr 2014, at 14:46, *sm1Ly <st.uz...@gmail.com> wrote:
> 
>> re2all. I use centos 6.5. this is my ceph.conf http://pastebin.com/0UCevzF5  
>> I doing manual deploying with this 
>> guidehttp://ceph.com/docs/master/install/manual-deployment/ when I get to 
>> step 15 I cant start my node, cause it doesnt defined in my conf. like this: 
>> /etc/init.d/ceph start mon.nfs2.abboom.world /etc/init.d/ceph: 
>> mon.nfs2.abboom.world not found (/etc/ceph/ceph.conf defines , /var/lib/ceph 
>> defines )  what I m doing wrong?
>> 
>> 
>> 
>> -- 
>> yours respectfully, Alexander Vasin.
>> 
>> 8 926 1437200
>> icq: 9906064
>> _______________________________________________
>> ceph-users mailing list
>> ceph-users@lists.ceph.com
>> http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
> 
> 
> 
> 
> -- 
> yours respectfully, Alexander Vasin.
> 
> 8 926 1437200
> icq: 9906064

_______________________________________________
ceph-users mailing list
ceph-users@lists.ceph.com
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com

Reply via email to