[zfs-discuss] smbfs and iscsi compile errors

Zenny garbytrash at gmail.com
Sun Feb 19 14:45:24 EST 2012

On 2/19/12, Steven Anderson <hikenboot at yahoo.com> wrote:
> Thanks Once Again turbo for that updated information!
> 1)  I have updated the document to include the New Git share and have added
> the information from Turbo's posting in the proper location of the document.
> 2)  Since I have no more KVM and Proxmox 2.0 installation running. I can not
> test. I am looking for volunteers to try out this configuration which should
> work, since I built mine step to step using these directions.
> http://wanfuse.blogspot.com/2011/12/debian-squeezezfsproxmox-20-kvm.html

It does not work. If you use the branch you specified for spl, it
fails when arc.o gets compiled during zfs comilation. Else the latest
spl branch makes the zfs comilation from turbo's repository fails even
after removing the suffix of the Release line in META.

And it is not possible to install from the ubuntu repository (though
https://launchpad.net/~zfs-native/+archive/stable states that lucid is
compatible with squeeze), but zfs-initramfs fails.

So in no way it is possible to compile with upstream kernel with
squeeze nor with pve. fyi.

> _______________________________________________________________
> another document/video people might be interested in is found here for those
> windows server guys running a mixed environment.
> http://wanfuse.blogspot.com/2012/01/installing-microsoft-sql-cluster-server.html
> <<--videos
> and here
> http://wanfuse.blogspot.com/2012/01/how-to-install-microsoft-mcsc-cluster.html
> <<---text version
> _________________________________________________________________
> --- On Sun, 2/19/12, Turbo Fredriksson <turbo at bayour.com> wrote:
> From: Turbo Fredriksson <turbo at bayour.com>
> Subject: Re: [zfs-discuss] smbfs and iscsi compile errors
> To: zfs-discuss at zfsonlinux.org
> Date: Sunday, February 19, 2012, 8:31 AM
> On Feb 19, 2012, at 5:18 PM, Steven Anderson wrote:
> I meant no offense to you. Please accept my apology!
> No worries, I did not take offence. I'm just stating the facts...
> You have indicated that that ISCSI Share command is all that's necessary not
> editing the following file: ietd.conf. Is this the correct code that I
> should leave out?
> Correct. This is done with the 'zfs set shareiscsi=on'. It actually runs a
> command ('ietadm') that
> creates the share, without editing the (that) config file.
> The actual command(s) are (see lib/libshare/README_iscsi.txt for more info):
>   /usr/sbin/ietadm --op new --tid 1 --params
> Name=iqn.2012-01.com.bayour:share.test1
>   /usr/sbin/ietadm --op new --tid 1 --lun 0 --params
> Path=/dev/zvol/share/test1,Type=fileio
> Here the volume is 'share/test1'. The TID, LUN and the IQN is calculated by
> the module.
> If you like to add optional parameters ('man ietadm'), you can create the
> script '/sbin/zfs_share_iscsi.sh'
> which will be called with only one parameter. The TID created by the ZFS
> module (in the commands
> above, this is '1' but will be increased for each 'share'). This can be
> shell, perl, python, whatever
> you like. As long as it accept only one parameter: the TID.
> Mine looks like this (as an example). In the future, this might instead be
> set to the 'shareiscsi'
> variable (instead of 'on' or 'off'). But I'd like the patch to mature a
> little more before i start on that.
> ----- s n i p -----celia:~# cat /sbin/zfs_share_iscsi.sh #!/bin/sh
> if [ -z "$1" ]; then    echo "Usage: `basename $0` [tid]"    exit 1fi
> TID=$1
> ietadm --op update --tid $TID --params
> MaxRecvDataSegmentLength=262144,MaxXmitDataSegmentLength=65536----- s n i p
> -----
> PS. Make sure it's executable! :)
> nano /etc/iet/ietd.conf
> [...]
> Correct, that's the one.

More information about the zfs-discuss mailing list