All of lore.kernel.org
 help / color / mirror / Atom feed
From: Wido den Hollander <wido@42on.com>
To: Dennis Chen <xschen@tnsoft.com.cn>
Cc: ceph-devel@vger.kernel.org
Subject: Re: question about ceph and FC SAN LUN data IO path
Date: Mon, 13 May 2013 14:34:11 +0200	[thread overview]
Message-ID: <5190DDC3.7030604@42on.com> (raw)
In-Reply-To: <5190DB83.9010007@tnsoft.com.cn>

On 05/13/2013 02:24 PM, Dennis Chen wrote:
> Hi,
>
> If I have an OSD attached by a FC SAN device with the connection like this:
> OSD HBA-->FC Switch-->SAN storage device.
>
> Then Map a LUN in the SAN to the OSD as its only disk device to save
> data. if I have a client also has an HBA and the topology diagram looks
> like:
>
>                    osd                                         client
>          -------        -------                            ------- -------
>         | HBA |      | NIC |  <---------------> | NIC |   | HBA |
>          ---+---      --------                            ------- -------
> \                                                            /
>                    \ /
>                         \                                   /
>                                \    --------------
>                                    | FC Switch |
>                                     --------------
>                                            |
>                                         SAN
>
> Suppose the created LUN above for OSD can be accessed by both osd and
> client though FC I/F, the question is:
> if I create a rbd block device in osd and map it to client, make a file
> system and mount it in client side, then create a file in the mounted
> FS. The read/write operation followed for this file will transfer data
> though ethernet path or FC path?
>

That will follow the Ethernet path.

I'm also trying to figure out why you would Ceph here and use RBD? If 
the client has direct access to the FC, why not have the client access 
that direcly?

Ceph makes sense when you are running it on multiple fully separated 
machines, but not if you run it all one one FC SAN imho.

Wido

> BRs,
> Dennis
>
> --
> 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


-- 
Wido den Hollander
42on B.V.

Phone: +31 (0)20 700 9902
Skype: contact42on

      reply	other threads:[~2013-05-13 12:34 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-05-13 12:24 question about ceph and FC SAN LUN data IO path Dennis Chen
2013-05-13 12:34 ` Wido den Hollander [this message]

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=5190DDC3.7030604@42on.com \
    --to=wido@42on.com \
    --cc=ceph-devel@vger.kernel.org \
    --cc=xschen@tnsoft.com.cn \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.