Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 01 Nov 2011 23:23:41 +0100
From:      Ivan Voras <ivoras@freebsd.org>
To:        freebsd-geom@freebsd.org
Cc:        freebsd-current@freebsd.org
Subject:   Re: RFC: GEOM MULTIPATH rewrite
Message-ID:  <j8prhs$tjg$1@dough.gmane.org>
In-Reply-To: <4EAF00A6.5060903@FreeBSD.org>
References:  <4EAF00A6.5060903@FreeBSD.org>

next in thread | previous in thread | raw e-mail | index | archive | help
On 31.10.2011. 21:10, Alexander Motin wrote:
> Hi.
>
> Attempt to fix some GEOM MULTIPATH issues made me almost rewrite it. So
> I would like to present my results and request for testing and feedback.

Good ideas, gmultipath robustness is a very good thing. Active/active 
support is also useful if ordering issues can be solved. There's a Linux 
paper linked at

http://en.wikipedia.org/wiki/Multipath_I/O

... but apparently it doesn't say much about this problem. Could a 
simple policy such as "block all IO iff an BIO_ORDERED request is being 
processed" work well enough?

I'd only add that you should be careful to be compatible with old 
gmultipath metadata (see e.g. recent user reports with 9.0 auto-updating 
gmirror metadata).


> The main changes:
>   - Improved locking and destruction process to fix crashes in many cases.
>   - Improved "automatic" configuration method to make it safe by reading
> metadata back from all specified paths after writing to one.
>   - Added provider size check to reduce chance of conflict with other
> GEOM classes.
>   - Added "manual" configuration method without using on-disk metadata.
>   - Added "add" and "remove" commands to manage paths manually.
>   - Failed paths no longer dropped from GEOM, but only marked as FAIL and
> excluded from I/O operations.
>   - Automatically restore failed paths when all others paths are marked
> as failed, for example, because of device-caused (not transport) errors.
>   - Added "fail" and "restore" commands to manually control FAIL flag.
>   - GEOM is now destroyed on last provider disconnection. IMHO it is
> right to do if device was completely removed.
>   - Added optional Active/Active mode support. Unlike Active/Passive
> mode, load evenly distributed between all working paths. If supported by
> device, it allows to significantly improve performance, utilizing
> bandwidth of all paths. It is controlled by -A option during creation.
> Disabled by default now.
>   - Improved `status` and `list` commands output.
>
> Latest patch can be found here:
> http://people.freebsd.org/~mav/gmultipath4.patch
>
> Feedbacks are welcome!
>
> Sponsored by: iXsystems, Inc.
>





Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?j8prhs$tjg$1>