https://bugzilla.redhat.com/show_bug.cgi?id=1478201 We have a libguestfs test which adds 256 virtio-scsi disks to a qemu virtual machine. The VM has 500 MB of RAM, 1 vCPU and no swap. This test has been failing for a little while. It runs out of memory during SCSI enumeration in early boot. Tonight I bisected the cause to: 5c279bd9e40624f4ab6e688671026d6005b066fa is the first bad commit commit 5c279bd9e40624f4ab6e688671026d6005b066fa Author: Christoph Hellwig <hch@xxxxxx> Date: Fri Jun 16 10:27:55 2017 +0200 scsi: default to scsi-mq Remove the SCSI_MQ_DEFAULT config option and default to the blk-mq I/O path now that we had plenty of testing, and have I/O schedulers for blk-mq. The module option to disable the blk-mq path is kept around for now. Signed-off-by: Christoph Hellwig <hch@xxxxxx> Signed-off-by: Martin K. Petersen <martin.petersen@xxxxxxxxxx> :040000 040000 57ec7d5d2ba76592a695f533a69f747700c31966 c79f6ecb070acc4fadf6fc05ca9ba32bc9c0c665 M drivers I also wrote a small test to see the maximum number of virtio-scsi disks I could add to the above VM. The results were very surprising (to me anyhow): With scsi-mq enabled: 175 disks With scsi-mq disabled: 1755 disks I don't know why the ratio is almost exactly 10 times. I read your slides about scsi-mq and it seems like a significant benefit to large machines, but could the out of the box defaults be made more friendly for small memory machines? Thanks, Rich. -- Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones Read my programming and virtualization blog: http://rwmj.wordpress.com libguestfs lets you edit virtual machines. Supports shell scripting, bindings from many languages. http://libguestfs.org