qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Qemu-devel] [PATCH] Add iSCSI support for QEMU


From: Ronnie Sahlberg
Subject: [Qemu-devel] [PATCH] Add iSCSI support for QEMU
Date: Wed, 21 Sep 2011 19:37:54 +1000


List,

Please find a patch that adds built-in iSCSI support to QEMU when built and 
linked against the multiplatform iscsi initiator library at  
git://github.com/sahlberg/libiscsi.git


I have implemented all suggested changed from Stefans review, except
* Fail the build if qemu blocksize is not multiple of 512.
Maybe QEMU will never use such, but if it does I dont think a blockdriver that 
can not support such should fail the build. Hence it just refuses to use the 
driver at runtime instead.

* I did not implement a synchronous bdrv_flush since I understand that Stefans 
patch will be merged shortly since it is required for other backend drivers 
anyway.


In some situations, using a built-in iscsi inititator has benefits against 
mounting the LUNs on the host.

* Privacy: The iSCSI LUNs are private to the guest and are not visible either 
to the host, nor to any processes running on the host.
* Ease of managment : If you have very many guests and very many, thousands of, 
iSCSI LUNs. It is inconvenient to have to expose all LUNs to the underlying 
host.
* No root requirement. Since the iSCSI LUNs are not mounted as devices on the 
host, ordinary users can set up and use iSCSI resources without the need for 
root privilege on the host to map the devices to local scsi devices.
* Since the network I/O for the iSCSI LUN is now done by the QEMU process 
instead of the kernel, each VM's iSCSI I/O traffic can be insolated & 
controlled via the cgroups network contorller / tc network classifier.
* Portable across OS. Each OS has different tools for setting up iSCSI usage. A 
native driver lets QEMU users setup iSCSI in the same way regardless of the 
level of OS support for iSCSI.

Please merge this patch to master or explain how I should change the patch so 
that it becomes acceptable for inclusion into QEMU.


regards
ronnie sahlberg




reply via email to

[Prev in Thread] Current Thread [Next in Thread]