From mboxrd@z Thu Jan 1 00:00:00 1970 From: "hzwulibin-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org" Subject: Does SSD Journal improve the performance? Date: Thu, 15 Oct 2015 09:29:03 +0800 Message-ID: <201510150929011726350@gmail.com> Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="===============0820178383==" Return-path: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: ceph-users-bounces-idqoXFIVOFJgJs9I8MT0rw@public.gmane.org Sender: "ceph-users" To: "ceph-devel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org" , ceph-users List-Id: ceph-devel.vger.kernel.org This is a multi-part message in MIME format. --===============0820178383== Content-Type: multipart/alternative; boundary="----=_001_NextPart877107666848_=----" This is a multi-part message in MIME format. ------=_001_NextPart877107666848_=---- Content-Type: text/plain; charset="GB2312" Content-Transfer-Encoding: base64 SGksIA0KDQpJdCBzaG91bGQgYmUgc3VyZSBTU0QgSm91cm5hbCB3aWxsIGltcHJvdmUgdGhlIHBl cmZvcm1hbmNlIG9mIElPUFMuIEJ1dCB1bmZvcnR1bmF0ZWx5IGl0J3Mgbm90IGluIG15IHRlc3Qu DQoNCkkgaGF2ZSB0d28gcG9vbHMgd2l0aCB0aGUgc2FtZSBudW1iZXIgb2Ygb3NkczoNCnBvb2wx LCBzc2RqX3NhczoNCjkgb3NkIHNlcnZlcnMsIDggT1NEcyhTQVMpIG9uIGV2ZXJ5IHNlcnZlcg0K Sm91cm5hbCBvbiBTU0QsIG9uZSBTU0QgZGlzayBmb3IgNCBTQVMgZGlza3MuDQoNCnBvb2wgMiwg c2FzOg0KOSBvc2Qgc2VydmVycywgOCBPU0RzKFNBUykgb24gZXZlcnkgc2VydmVyDQpKb3VybmFs IG9uIFNBUyBkaXNrIGl0c2VsZqGjDQoNCkkgdXNlIHJiZCB0byBjcmVhdGUgYSB2b2x1bWUgaW4g cG9vbDEgYW5kIHBvb2wyIHNlcGFyYXRlbHkgYW5kIHVzZSBmaW8gdG8gdGVzdA0KdGhlIHJhbmQg d3JpdGUgSU9QU6GjaGVyZSBpcyB0aGUgZmlvIGNvbmZpZ3VyYXRpb246DQoNCnJ3PXJhbmR3cml0 ZQ0KaW9lbmdpbmU9bGliYWlvDQpkaXJlY3Q9MQ0KaW9kZXB0aD0xMjgNCmJzPTRrDQpudW1qb2Jz PTENCg0KVGhlIHJlc3VsdCBpIGdvdCBpczoNCnZvbHVtZSBpbiBwb29sMSwgYWJvdXQgNWsNCnZv bHVtZSBpbiBwb29sMiwgYWJvdXQgMTJrDQoNCkl0J3MgYSBiaWcgZ2FwIGhlcmUsIGFueW9uZSBj YW4gZ2l2ZSBtZSBzb21lIHN1Z2dlc3Rpb24gaGVyZT8NCg0KY2VwaCB2ZXJzaW9uOiBoYW1tZXIo MC45NC4zKQ0Ka2VybmVsOiAzLjEwDQoNCg0KDQpoend1bGliaW5AZ21haWwuY29tDQo= ------=_001_NextPart877107666848_=---- Content-Type: text/html; charset="GB2312" Content-Transfer-Encoding: quoted-printable =0A
= Hi, 

It should be sure SSD Journal will improve the performance of IO= PS. But unfortunately it's not in my test.

I have two pools with the same n= umber of osds:
pool1, ssdj_sas:
9 osd servers, 8 OSDs(SAS) on every server
=
Journal on SSD, one SSD disk for 4 SAS disks.=

= pool 2, sas:
9 osd servers, 8 OSDs(= SAS) on every server
Journal on SAS disk itself=A1=A3

I use= rbd to create a volume in pool1 and pool2 separately and use fio to test<= /div>
the rand write IOPS=A1=A3here is the fio= configuration:

rw=3Drandwrite
ioengine= =3Dlibaio
direct=3D1
iodepth=3D128
bs=3D4k
numjobs=3D1

The result i got is:
volume in pool1, about 5k
volume in pool2, about 12k

=
It's a big gap here, anyone can give me= some suggestion here?

ceph version: hammer(0.94.3)
kernel: 3.10
=0A

=0A
hzwulibin-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org
=0A ------=_001_NextPart877107666848_=------ --===============0820178383== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ ceph-users mailing list ceph-users-idqoXFIVOFJgJs9I8MT0rw@public.gmane.org http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com --===============0820178383==--