linux-lvm.redhat.com archive mirror
 help / color / mirror / Atom feed
From: Ken Goods <kgoods@cropusainsurance.com>
To: Peter Rajnoha <prajnoha@redhat.com>
Cc: linux-lvm Mailing List <linux-lvm@redhat.com>
Subject: Re: [linux-lvm] multipath_component_detection
Date: Tue, 17 Jan 2017 08:27:40 -0800 (PST)	[thread overview]
Message-ID: <15337358.38.1484670456030.JavaMail.kgoods@Ken-PC> (raw)
In-Reply-To: <afb0ab23-80bd-97bc-4e87-0e4e65e5196e@redhat.com>

[-- Attachment #1: Type: text/plain, Size: 2456 bytes --]

Thanks so much Peter, exactly what I was looking for. Also thanks for the heads up on external_device_info_source, I'll use it. 

----- Original Message -----

From: "Peter Rajnoha" <prajnoha@redhat.com> 
To: kgoods@cropusainsurance.com 
Cc: "LVM general discussion and development" <linux-lvm@redhat.com> 
Sent: Tuesday, January 17, 2017 3:19:44 AM 
Subject: Re: [linux-lvm] multipath_component_detection 

On 01/16/2017 08:46 PM, Ken Goods wrote: 
> Hi all, 
> Just a quick question. I've been setting up a (iscsi) LUN and want to 
> create an LVM PV on it. 
> 
> The docs I have read are somewhat conflicted. Some say that the 
> underlying devices should be filtered, however according to comments in 
> lvm.conf it says that if multipath_component_detection is enabled, "LVM2 
> will ignore devices used as component paths of device-mapper multipath 
> devices". 
> 
> So filtering is no longer necessary with LVM2? 
> 

The multipath component detection is an automatic filter in LVM2 (if 
enabled with devices/multipath_component_detection=1 in LVM 
configuration - that is used by default). So in this case you don't need 
to set up devices/global_filter (or devices/filter) manually. LVM will 
filter multipath components automatically by looking at the device stack 
and if it detects that there's a multipath device on top of certain set 
of devices, such devices are assumed as multipath components and they're 
filtered automatically. 

Of course, for this automatic filter to work correctly, you also need to 
have your multipath correctly configured - LVM doesn't know that the 
devices are multipath components unless there's multipath device already 
set up and running on top of those components. 

However, there's one improvement in this detection if you use 
devices/external_device_info_source="udev" in your LVM configuration 
(this is not used by default yet). In this case, LVM reads information 
from udev database which in turn has this information about multipath 
components directly from multipath and that one takes this information 
from its own configuration (the component WWID list in multipath's 
configuraton) - so in this case you don't even need to have multipath 
device set up yet on top of multipath components for LVM to detect 
multipath components properly. 

But usually, if your system is properly configured, you always have 
multipath device set up and running on top of multipath components. 

-- 
Peter 


[-- Attachment #2: Type: text/html, Size: 2980 bytes --]

      reply	other threads:[~2017-01-17 16:27 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <11587299.215.1484595906899.JavaMail.kgoods@Ken-PC>
2017-01-16 19:46 ` [linux-lvm] multipath_component_detection Ken Goods
2017-01-17 11:19   ` Peter Rajnoha
2017-01-17 16:27     ` Ken Goods [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=15337358.38.1484670456030.JavaMail.kgoods@Ken-PC \
    --to=kgoods@cropusainsurance.com \
    --cc=linux-lvm@redhat.com \
    --cc=prajnoha@redhat.com \
    /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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).