linux-kernel - md new raid read balance, i need help to end this new feature

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [day] [month] [year] [list]

Message-ID: <BAY144-W17A513D39A8C38BA849520BED20@phx.gbl>
Date:	Wed, 16 Feb 2011 00:56:22 -0200
From:	Roberto Spadim <robertospadim@...mail.com>
To:	<linux-kernel@...r.kernel.org>
Subject: md new raid read balance, i need help to end this new feature -
 queue block informations

hi guys, i´m making a new read_balance function for md (linux software raid) can anyone help me?
i need informations about queue of each mirror (/dev/sda, /dev/nbd, /dev/hda, /dev/XXX = block devices)
what information?
estimate time to end queue commands

why? with estimate time i can select the best mirror (raid1) to read, the time with less queue time + less read time, is the best disk to read

i know there´s some elevators and some devices that don´t work with queue, or have user space queue that is dificult to get estimate value
in this case i want:
total sectors to write (sequencial/non sequencial if possible)
total sectors to read (sequencial/non sequencial if possible)
first and last sector number (to get disk position change time, i´m considering that elevator get the smallest patch, so the last-first is the head distance, head distance * distance/sector give me a good estimation value)

i think that´s all

in raid1.c i have the generic_make_request() call, can i get all informations with those parameters?

sorry i´m from brazil if i wrote something out of order, or difficult to understand please report it.
the today implementation is kernel 2.6.37 based, source at:
www.spadim.com.br/raid1/

any help is welcome
thanks
 		 	   		  --
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/