Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 14 Apr 2013 12:01:43 +0200
From:      mxb <mxb@alumni.chalmers.se>
To:        "freebsd-fs@freebsd.org" <freebsd-fs@freebsd.org>
Subject:   Re: ZFS: ZIL device export/import
Message-ID:  <9EE9328B-40B1-4510-B404-242D0F2C7697@alumni.chalmers.se>
In-Reply-To: <op.wvhyvkzx8527sy@pinky>
References:  <5A2824CA-2A67-47FA-AB27-20C6EBD2C501@alumni.chalmers.se> <51699B8E.7050003@platinum.linux.pl> <BCBD7CDE-1BBB-4855-9240-897770FEF822@alumni.chalmers.se> <op.wvhrfrhj8527sy@pinky> <2DE8AD5E-B84C-4D88-A242-EA30EA4A68FD@alumni.chalmers.se> <op.wvhyvkzx8527sy@pinky>

next in thread | previous in thread | raw e-mail | index | archive | help

Well, I'm trying to preclude any undesired effect in the whole setup, as =
this is going to production.

SAS-link might not be a bottleneck here and I'm overreacting.

Locally ,on per HU basis, I have 6Gbit/s SAS/SATA. Both card and =
disks(SSD) attached to it.
SAS Expander is also 6Gbit/s, attaching 10k RPM SAS mechanical disks on =
JBOD.

I use Intel 520 SSD and Pulsar SSD in this setup.
ZIL resided locally on Intel SSD(per HU), but now will probably move to =
Pulsar SSD(moved to JBOD as those disks have dual SAS/SATA link). L2ARC =
resided on Pulsar (Pulsar was in each HU. eg. I have 2x Pulsar).

Looks like I have to re-design the whole setup, as of ZIL.

//mxb


On 13 apr 2013, at 22:51, Ronald Klop <ronald-freebsd8@klop.yi.org> =
wrote:

> I thought the idea of ZIL is a fast buffer before the write to slow =
disk. Are you really sure the SAS expander is the bottleneck in the =
system instead of the disks?




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?9EE9328B-40B1-4510-B404-242D0F2C7697>