OSD per disk.

OSD per disk.

[chandrashekhar]

Hi,

      I want to setup a ceph cluster with 4 osds ( one OSD per disk), 2 mons, 1
mds. Please help me to configure this.

I have a machine with following hardware:

4 hard disks ( each 500GB).

/dev/sda - 500G ( OS )  
/dev/sdb - 500G
/dev/sdc - 500G
/dev/sdd - 500G


OS: ubuntu 12.O4 server ( 64 bit )




Thanks,
Chandra.

Re: OSD per disk.

[Alexandre DERUMIER]

Hi, just defined osds in ceph.conf, they mount each disk in correct mountpoint.

this example : 
1st server :
  /dev/sdb1 -> /data/osd0
  /dev/sdc1 -> /data/osd1
  /dev/sdd1 -> /data/osd2
  ....
same for journal.(you can use differents disk for journal or use another partition on same disk)



ceph.conf
[osd] 

osd data = /data/osd$id 
osd journal = /data/osd$id/journal 
osd journal size = 512 

[osd.0] 
host = ceph-node-1 


[osd.1] 
host = ceph-node-1 


[osd.2] 
host = ceph-node-1 



[osd.3] 
host = ceph-node-2 


[osd.4] 
host = ceph-node-2 


----- Mail original ----- 

De: "chandrashekhar" <chandubigc@gmail.com> 
À: ceph-devel@vger.kernel.org 
Envoyé: Lundi 28 Mai 2012 08:19:17 
Objet: OSD per disk. 

Hi, 

I want to setup a ceph cluster with 4 osds ( one OSD per disk), 2 mons, 1 
mds. Please help me to configure this. 

I have a machine with following hardware: 

4 hard disks ( each 500GB). 

/dev/sda - 500G ( OS ) 
/dev/sdb - 500G 
/dev/sdc - 500G 
/dev/sdd - 500G 


OS: ubuntu 12.O4 server ( 64 bit ) 




Thanks, 
Chandra. 

-- 
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in 
the body of a message to majordomo@vger.kernel.org 
More majordomo info at http://vger.kernel.org/majordomo-info.html 



-- 

-- 




	Alexandre D erumier 
Ingénieur Système 
Fixe : 03 20 68 88 90 
Fax : 03 20 68 90 81 
45 Bvd du Général Leclerc 59100 Roubaix - France 
12 rue Marivaux 75002 Paris - France 
	
--
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: OSD per disk.

[chandrashekhar]

Alexandre DERUMIER <aderumier <at> odiso.com> writes:

> 
> Hi, just defined osds in ceph.conf, they mount each disk in correct mountpoint.
> 
> this example : 
> 1st server :
>   /dev/sdb1 -> /data/osd0
>   /dev/sdc1 -> /data/osd1
>   /dev/sdd1 -> /data/osd2
>   ....
> same for journal.(you can use differents disk for journal or use another
partition on same disk)
> 
> ceph.conf
> [osd] 
> 
> osd data = /data/osd$id 
> osd journal = /data/osd$id/journal 
> osd journal size = 512 
> 
> [osd.0] 
> host = ceph-node-1 
> 
> [osd.1] 
> host = ceph-node-1 
> 
> [osd.2] 
> host = ceph-node-1 
> 
> [osd.3] 
> host = ceph-node-2 
> 
> [osd.4] 
> host = ceph-node-2 
> 
> ----- Mail original ----- 
> 
> De: "chandrashekhar" <chandubigc <at> gmail.com> 
> À: ceph-devel <at> vger.kernel.org 
> Envoyé: Lundi 28 Mai 2012 08:19:17 
> Objet: OSD per disk. 
> 
> Hi, 
> 
> I want to setup a ceph cluster with 4 osds ( one OSD per disk), 2 mons, 1 
> mds. Please help me to configure this. 
> 
> I have a machine with following hardware: 
> 
> 4 hard disks ( each 500GB). 
> 
> /dev/sda - 500G ( OS ) 
> /dev/sdb - 500G 
> /dev/sdc - 500G 
> /dev/sdd - 500G 
> 
> OS: ubuntu 12.O4 server ( 64 bit ) 
> 
> Thanks, 
> Chandra. 
> 

Thanks Alexandre,  I created four directories in /data  (osd0,osd1,osd2,osd3) 
and mounted as below:

/dev/sdb1 -> /data/osd1
/dev/sdc1 -> /data/osd2
/dev/sdd1 -> /data/osd3


But when I start ceph its starting mons and md daemons but not osds. Please help
me to get this working.

>ceph -w
2012-05-28 14:43:23.756895    pg v2: 792 pgs: 792 creating; 0 bytes data, 0 KB
used, 0 KB / 0 KB avail
2012-05-28 14:43:23.762160   mds e1: 0/0/1 up
2012-05-28 14:43:23.762296   osd e1: 0 osds: 0 up, 0 in
2012-05-28 14:43:23.762586   log 2012-05-28 14:24:15.391671 mon.0
192.168.7.10:6789/0 4 : [INF] mon.a@0 won leader election with quorum 0,1
2012-05-28 14:43:23.762741   mon e1: 2 mons at
{a=192.168.7.10:6789/0,b=192.168.7.10:6790/0}




Here is my ceph.conf:
-------------------------------------------------
[global]
   auth supported = cephx
   keyring = /etc/ceph/keyring.admin

[mon]
   mon data = /data/mon.$id

[mds]

[osd]
   osd data = /data/osd$id
   osd journal = /data/osd$id/osd.$id.journal
   osd journal size = 1000



[mon.a]
   host = testnode
   mon addr = 192.168.7.10:6789
[mon.b]
  host = testnode
  mon addr = 192.168.7.10:6790

[osd.0]
   host = testnode

[osd.1]
   host = testnode

[osd.2]
   host = testnode

[osd.3]
   host = testnode
[mds.a]
   host = testnode




--
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: OSD per disk.

[Alexandre DERUMIER]

maybe try 

[osd.0] 
host = testnode 
osd data = /data/osd0 
osd journal = /data/osd0/osd0journal 
osd journal size = 1000 

[osd.1] 
host = testnode 
osd data = /data/osd1 
osd journal = /data/osd1/osd1journal 
osd journal size = 1000 
[osd.2] 
host = testnode 
osd data = /data/osd2 
osd journal = /data/osd2/osd2journal 
osd journal size = 1000 
[osd.3] 
host = testnode 
osd data = /data/osd3 
osd journal = /data/osd3/osd3journal 
osd journal size = 1000 



you can start osd with

/etc/init.d/ceph start osd0
/etc/init.d/ceph start osd1
/etc/init.d/ceph start osd2
/etc/init.d/ceph start osd3

----- Mail original ----- 

De: "chandrashekhar" <chandubigc@gmail.com> 
À: ceph-devel@vger.kernel.org 
Envoyé: Lundi 28 Mai 2012 11:25:36 
Objet: Re: OSD per disk. 


Alexandre DERUMIER <aderumier <at> odiso.com> writes: 

> 
> Hi, just defined osds in ceph.conf, they mount each disk in correct mountpoint. 
> 
> this example : 
> 1st server : 
> /dev/sdb1 -> /data/osd0 
> /dev/sdc1 -> /data/osd1 
> /dev/sdd1 -> /data/osd2 
> .... 
> same for journal.(you can use differents disk for journal or use another 
partition on same disk) 
> 
> ceph.conf 
> [osd] 
> 
> osd data = /data/osd$id 
> osd journal = /data/osd$id/journal 
> osd journal size = 512 
> 
> [osd.0] 
> host = ceph-node-1 
> 
> [osd.1] 
> host = ceph-node-1 
> 
> [osd.2] 
> host = ceph-node-1 
> 
> [osd.3] 
> host = ceph-node-2 
> 
> [osd.4] 
> host = ceph-node-2 
> 
> ----- Mail original ----- 
> 
> De: "chandrashekhar" <chandubigc <at> gmail.com> 
> À: ceph-devel <at> vger.kernel.org 
> Envoyé: Lundi 28 Mai 2012 08:19:17 
> Objet: OSD per disk. 
> 
> Hi, 
> 
> I want to setup a ceph cluster with 4 osds ( one OSD per disk), 2 mons, 1 
> mds. Please help me to configure this. 
> 
> I have a machine with following hardware: 
> 
> 4 hard disks ( each 500GB). 
> 
> /dev/sda - 500G ( OS ) 
> /dev/sdb - 500G 
> /dev/sdc - 500G 
> /dev/sdd - 500G 
> 
> OS: ubuntu 12.O4 server ( 64 bit ) 
> 
> Thanks, 
> Chandra. 
> 

Thanks Alexandre, I created four directories in /data (osd0,osd1,osd2,osd3) 
and mounted as below: 

/dev/sdb1 -> /data/osd1 
/dev/sdc1 -> /data/osd2 
/dev/sdd1 -> /data/osd3 


But when I start ceph its starting mons and md daemons but not osds. Please help 
me to get this working. 

>ceph -w 
2012-05-28 14:43:23.756895 pg v2: 792 pgs: 792 creating; 0 bytes data, 0 KB 
used, 0 KB / 0 KB avail 
2012-05-28 14:43:23.762160 mds e1: 0/0/1 up 
2012-05-28 14:43:23.762296 osd e1: 0 osds: 0 up, 0 in 
2012-05-28 14:43:23.762586 log 2012-05-28 14:24:15.391671 mon.0 
192.168.7.10:6789/0 4 : [INF] mon.a@0 won leader election with quorum 0,1 
2012-05-28 14:43:23.762741 mon e1: 2 mons at 
{a=192.168.7.10:6789/0,b=192.168.7.10:6790/0} 




Here is my ceph.conf: 
------------------------------------------------- 
[global] 
auth supported = cephx 
keyring = /etc/ceph/keyring.admin 

[mon] 
mon data = /data/mon.$id 

[mds] 

[osd] 
osd data = /data/osd$id 
osd journal = /data/osd$id/osd.$id.journal 
osd journal size = 1000 



[mon.a] 
host = testnode 
mon addr = 192.168.7.10:6789 
[mon.b] 
host = testnode 
mon addr = 192.168.7.10:6790 

[osd.0] 
host = testnode 

[osd.1] 
host = testnode 

[osd.2] 
host = testnode 

[osd.3] 
host = testnode 
[mds.a] 
host = testnode 




-- 
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in 
the body of a message to majordomo@vger.kernel.org 
More majordomo info at http://vger.kernel.org/majordomo-info.html 



-- 

-- 




	Alexandre D erumier 
Ingénieur Système 
Fixe : 03 20 68 88 90 
Fax : 03 20 68 90 81 
45 Bvd du Général Leclerc 59100 Roubaix - France 
12 rue Marivaux 75002 Paris - France 
	
--
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: OSD per disk.

[Tommi Virtanen]

On Mon, May 28, 2012 at 2:34 AM, Alexandre DERUMIER <aderumier@odiso.com> wrote:
> maybe try
>
> [osd.0]
> host = testnode
> osd data = /data/osd0
> osd journal = /data/osd0/osd0journal
> osd journal size = 1000

That shouldn't be needed. osd.0 will happily read the [osd] section in
the config file.

Re: OSD per disk.

[Tommi Virtanen]

On Mon, May 28, 2012 at 2:25 AM, chandrashekhar <chandubigc@gmail.com> wrote:
> Thanks Alexandre,  I created four directories in /data  (osd0,osd1,osd2,osd3)
> and mounted as below:
>
> /dev/sdb1 -> /data/osd1
> /dev/sdc1 -> /data/osd2
> /dev/sdd1 -> /data/osd3
>
>
> But when I start ceph its starting mons and md daemons but not osds. Please help
> me to get this working.

How did you create the cluster? mkcephfs?

Do you see log entries in /var/log/ceph/*osd*.log ? What do they say?
--
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Fwd: OSD per disk.

[Tommi Virtanen]

For the benefit of the mailing list; there's an assert(nlock > 0)
crash in there.

---------- Forwarded message ----------
From: chandrashekhar mekala <chandubigc@gmail.com>
Date: Tue, May 29, 2012 at 9:51 PM
Subject: Re: OSD per disk.
To: Tommi Virtanen <tv@inktank.com>


Yes , I am using mkcephfs to create  cluster.

osd.1.log
-------------------------
2012-05-30 10:17:58.948426 7f98b4856780 journal close /data/osd.1/osd.1.journal
./common/Mutex.h: In function 'void Mutex::Unlock()' thread
7f98b4856780 time 2012-05-30 10:17:58.948800
./common/Mutex.h: 117: FAILED assert(nlock > 0)
 ceph version 0.41 (commit:c1345f7136a0af55d88280ffe4b58339aaf28c9d)
 1: (Mutex::Unlock()+0x93) [0x4f5c73]
 2: (OSD::init()+0x459) [0x546b79]
 3: (main()+0x2496) [0x4a8bc6]
 4: (__libc_start_main()+0xed) [0x7f98b2d2676d]
 5: /usr/bin/ceph-osd() [0x4aa68d]
 ceph version 0.41 (commit:c1345f7136a0af55d88280ffe4b58339aaf28c9d)
 1: (Mutex::Unlock()+0x93) [0x4f5c73]
 2: (OSD::init()+0x459) [0x546b79]
 3: (main()+0x2496) [0x4a8bc6]
 4: (__libc_start_main()+0xed) [0x7f98b2d2676d]
 5: /usr/bin/ceph-osd() [0x4aa68d]
*** Caught signal (Aborted) **
 in thread 7f98b4856780
 ceph version 0.41 (commit:c1345f7136a0af55d88280ffe4b58339aaf28c9d)
 1: /usr/bin/ceph-osd() [0x5fb1d6]
 2: (()+0xfcb0) [0x7f98b3ef2cb0]
 3: (gsignal()+0x35) [0x7f98b2d3b445]
 4: (abort()+0x17b) [0x7f98b2d3ebab]
 5: (__gnu_cxx::__verbose_terminate_handler()+0x11d) [0x7f98b368969d]
 6: (()+0xb5846) [0x7f98b3687846]
 7: (()+0xb5873) [0x7f98b3687873]
 8: (()+0xb596e) [0x7f98b368796e]
 9: (ceph::__ceph_assert_fail(char const*, char const*, int, char
const*)+0x200) [0x5ce010]
 10: (Mutex::Unlock()+0x93) [0x4f5c73]
 11: (OSD::init()+0x459) [0x546b79]
 12: (main()+0x2496) [0x4a8bc6]
 13: (__libc_start_main()+0xed) [0x7f98b2d2676d]
 14: /usr/bin/ceph-osd() [0x4aa68d]




On Tue, May 29, 2012 at 9:56 PM, Tommi Virtanen <tv@inktank.com> wrote:
>
> On Mon, May 28, 2012 at 2:25 AM, chandrashekhar <chandubigc@gmail.com> wrote:
> > Thanks Alexandre,  I created four directories in /data  (osd0,osd1,osd2,osd3)
> > and mounted as below:
> >
> > /dev/sdb1 -> /data/osd1
> > /dev/sdc1 -> /data/osd2
> > /dev/sdd1 -> /data/osd3
> >
> >
> > But when I start ceph its starting mons and md daemons but not osds. Please help
> > me to get this working.
>
> How did you create the cluster? mkcephfs?
>
> Do you see log entries in /var/log/ceph/*osd*.log ? What do they say?
--
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: OSD per disk.

[Sam Just]

That bug was fixed in f859f25d7eaec01fb6a82b88409d005c8177fea2.

It actually appears to mean that the osd failed to authenticate.
-Sam

On Wed, May 30, 2012 at 9:53 AM, Tommi Virtanen <tv@inktank.com> wrote:
> For the benefit of the mailing list; there's an assert(nlock > 0)
> crash in there.
>
> ---------- Forwarded message ----------
> From: chandrashekhar mekala <chandubigc@gmail.com>
> Date: Tue, May 29, 2012 at 9:51 PM
> Subject: Re: OSD per disk.
> To: Tommi Virtanen <tv@inktank.com>
>
>
> Yes , I am using mkcephfs to create  cluster.
>
> osd.1.log
> -------------------------
> 2012-05-30 10:17:58.948426 7f98b4856780 journal close /data/osd.1/osd.1.journal
> ./common/Mutex.h: In function 'void Mutex::Unlock()' thread
> 7f98b4856780 time 2012-05-30 10:17:58.948800
> ./common/Mutex.h: 117: FAILED assert(nlock > 0)
>  ceph version 0.41 (commit:c1345f7136a0af55d88280ffe4b58339aaf28c9d)
>  1: (Mutex::Unlock()+0x93) [0x4f5c73]
>  2: (OSD::init()+0x459) [0x546b79]
>  3: (main()+0x2496) [0x4a8bc6]
>  4: (__libc_start_main()+0xed) [0x7f98b2d2676d]
>  5: /usr/bin/ceph-osd() [0x4aa68d]
>  ceph version 0.41 (commit:c1345f7136a0af55d88280ffe4b58339aaf28c9d)
>  1: (Mutex::Unlock()+0x93) [0x4f5c73]
>  2: (OSD::init()+0x459) [0x546b79]
>  3: (main()+0x2496) [0x4a8bc6]
>  4: (__libc_start_main()+0xed) [0x7f98b2d2676d]
>  5: /usr/bin/ceph-osd() [0x4aa68d]
> *** Caught signal (Aborted) **
>  in thread 7f98b4856780
>  ceph version 0.41 (commit:c1345f7136a0af55d88280ffe4b58339aaf28c9d)
>  1: /usr/bin/ceph-osd() [0x5fb1d6]
>  2: (()+0xfcb0) [0x7f98b3ef2cb0]
>  3: (gsignal()+0x35) [0x7f98b2d3b445]
>  4: (abort()+0x17b) [0x7f98b2d3ebab]
>  5: (__gnu_cxx::__verbose_terminate_handler()+0x11d) [0x7f98b368969d]
>  6: (()+0xb5846) [0x7f98b3687846]
>  7: (()+0xb5873) [0x7f98b3687873]
>  8: (()+0xb596e) [0x7f98b368796e]
>  9: (ceph::__ceph_assert_fail(char const*, char const*, int, char
> const*)+0x200) [0x5ce010]
>  10: (Mutex::Unlock()+0x93) [0x4f5c73]
>  11: (OSD::init()+0x459) [0x546b79]
>  12: (main()+0x2496) [0x4a8bc6]
>  13: (__libc_start_main()+0xed) [0x7f98b2d2676d]
>  14: /usr/bin/ceph-osd() [0x4aa68d]
>
>
>
>
> On Tue, May 29, 2012 at 9:56 PM, Tommi Virtanen <tv@inktank.com> wrote:
>>
>> On Mon, May 28, 2012 at 2:25 AM, chandrashekhar <chandubigc@gmail.com> wrote:
>> > Thanks Alexandre,  I created four directories in /data  (osd0,osd1,osd2,osd3)
>> > and mounted as below:
>> >
>> > /dev/sdb1 -> /data/osd1
>> > /dev/sdc1 -> /data/osd2
>> > /dev/sdd1 -> /data/osd3
>> >
>> >
>> > But when I start ceph its starting mons and md daemons but not osds. Please help
>> > me to get this working.
>>
>> How did you create the cluster? mkcephfs?
>>
>> Do you see log entries in /var/log/ceph/*osd*.log ? What do they say?
--
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: OSD per disk.

[chandrashekhar]

>>
I built ceph from source ( f859f25d7eaec01fb6a82b88409d005c8177fea2  branch).

mkcephfs -a -c /etc/ceph/ceph.conf  -k /etc/ceph/keyring.admin -- is working
fine, but when I tried to start ceph its showing following error.

>strace ceph -a start

rt_sigprocmask(SIG_SETMASK, [PIPE], NULL, 8) = 0
rt_sigprocmask(SIG_BLOCK, [PIPE], [PIPE], 8) = 0
mmap(NULL, 8392704, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS|MAP_STACK,
-1, 0) = 0x7f1cc5a49000
mprotect(0x7f1cc5a49000, 4096, PROT_NONE) = 0
clone(child_stack=0x7f1cc6248f70,
flags=CLONE_VM|CLONE_FS|CLONE_FILES|CLONE_SIGHAND|CLONE_THREAD|CLONE_SYSVSEM
|CLONE_SETTLS|CLONE_PARENT_SETTID|CLONE_CHILD_CLEARTID,
parent_tidptr=0x7f1cc62499d0, tls=0x7f1cc6249700, 
child_tidptr=0x7f1cc62499d0) =
21042
rt_sigprocmask(SIG_SETMASK, [PIPE], NULL, 8) = 0
rt_sigprocmask(SIG_BLOCK, [PIPE], [PIPE], 8) = 0
mmap(NULL, 1052672, PROT_READ|PROT_WRITE, 
MAP_PRIVATE|MAP_ANONYMOUS|MAP_STACK,
-1, 0) = 0x7f1ccaa98000
mprotect(0x7f1ccaa98000, 4096, PROT_NONE) = 0
clone(child_stack=0x7f1ccab97f70,
flags=CLONE_VM|CLONE_FS|CLONE_FILES|CLONE_SIGHAND|CLONE_THREAD|
CLONE_SYSVSEM|CLONE_SETTLS|CLONE_PARENT_SETTID|CLONE_CHILD_CLEARTID,
parent_tidptr=0x7f1ccab989d0, tls=0x7f1ccab98700, child_tidptr=0x7f1ccab989d0) =
21043
rt_sigprocmask(SIG_SETMASK, [PIPE], NULL, 8) = 0
futex(0x201b904, FUTEX_WAIT_PRIVATE, 1, NULL

Re: OSD per disk.

[Tommi Virtanen]

On Tue, Jun 5, 2012 at 5:34 AM, chandrashekhar <chandubigc@gmail.com> wrote:
>>>
> I built ceph from source ( f859f25d7eaec01fb6a82b88409d005c8177fea2  branch).
>
> mkcephfs -a -c /etc/ceph/ceph.conf  -k /etc/ceph/keyring.admin -- is working
> fine, but when I tried to start ceph its showing following error.

I'm sorry, perhaps part of the email got cut or something, but all I
see is a snippet of strace where things succeed -- what's the error
you saw?

>>strace ceph -a start
>
> rt_sigprocmask(SIG_SETMASK, [PIPE], NULL, 8) = 0
> rt_sigprocmask(SIG_BLOCK, [PIPE], [PIPE], 8) = 0
> mmap(NULL, 8392704, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS|MAP_STACK,
> -1, 0) = 0x7f1cc5a49000
> mprotect(0x7f1cc5a49000, 4096, PROT_NONE) = 0
> clone(child_stack=0x7f1cc6248f70,
> flags=CLONE_VM|CLONE_FS|CLONE_FILES|CLONE_SIGHAND|CLONE_THREAD|CLONE_SYSVSEM
> |CLONE_SETTLS|CLONE_PARENT_SETTID|CLONE_CHILD_CLEARTID,
> parent_tidptr=0x7f1cc62499d0, tls=0x7f1cc6249700,
> child_tidptr=0x7f1cc62499d0) =
> 21042
> rt_sigprocmask(SIG_SETMASK, [PIPE], NULL, 8) = 0
> rt_sigprocmask(SIG_BLOCK, [PIPE], [PIPE], 8) = 0
> mmap(NULL, 1052672, PROT_READ|PROT_WRITE,
> MAP_PRIVATE|MAP_ANONYMOUS|MAP_STACK,
> -1, 0) = 0x7f1ccaa98000
> mprotect(0x7f1ccaa98000, 4096, PROT_NONE) = 0
> clone(child_stack=0x7f1ccab97f70,
> flags=CLONE_VM|CLONE_FS|CLONE_FILES|CLONE_SIGHAND|CLONE_THREAD|
> CLONE_SYSVSEM|CLONE_SETTLS|CLONE_PARENT_SETTID|CLONE_CHILD_CLEARTID,
> parent_tidptr=0x7f1ccab989d0, tls=0x7f1ccab98700, child_tidptr=0x7f1ccab989d0) =
> 21043
> rt_sigprocmask(SIG_SETMASK, [PIPE], NULL, 8) = 0
> futex(0x201b904, FUTEX_WAIT_PRIVATE, 1, NULL
--
To unsubscribe from this list: send the line "unsubscribe ceph-devel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: OSD per disk.

[David McBride]

On Mon, 2012-05-28 at 06:19 +0000, chandrashekhar wrote:

>       I want to setup a ceph cluster with 4 osds ( one OSD per disk), 2 mons, 1
> mds. Please help me to configure this.

You almost certainly do not want to have 2 MON nodes in your cluster; an
odd number of nodes is strongly recommended.

(Reason: a strict majority of MON odes must be up and functional to
maintain quorum; increasing the number of MONs from 1 to 2 makes
failures more likely but does not increase resiliency.)

Cheers,
David
-- 
David McBride <dwm@doc.ic.ac.uk>
Department of Computing, Imperial College, London