[Qemu-discuss] QEMU caching modes

2017-03-17 Thread Jan Schermer
Hi, I am investigating a weird caching behaviour of QEMU on Ubuntu 14.04 and most likely even 16.04. Our VMs are using DRBD devices as drives, like so: file=/dev/drbd/by-res/3fa0da3f-9607-4068-b517-349c05157d9a/0,format=raw,if=none,id=drive-virtio-disk0,serial=1f59fd16-72e6-45f8-b047-e8a876bee5f

Re: [Qemu-discuss] QEMU caching modes

2017-03-20 Thread Jan Schermer
Direct IO != Synchronous IO I'm seeing Synchronous IO when I shouldn't. Jan > On 20 Mar 2017, at 17:21, Alberto Garcia wrote: > > On Fri, Mar 17, 2017 at 09:52:27PM +0100, Jan Schermer wrote: > >> I'm trying to understand when/what IO is synchronous and why

Re: [Qemu-discuss] QEMU caching modes

2017-03-21 Thread Jan Schermer
27;s Linux > kernel? > dirty_expire_centisecs > dirty_writeback_centisecs > > El 20/03/17 a les 17:21, Alberto Garcia ha escrit: >> On Fri, Mar 17, 2017 at 09:52:27PM +0100, Jan Schermer wrote: >> >>> I'm trying to understand when/what IO is synchronous and why,

Re: [Qemu-discuss] Increasing number of threads when using ceph

2017-06-01 Thread Jan Schermer
My experience is the same, threads only grow. The only way to decrease the number of threads is to detach the block device completely. Note that if you need to support VMs with many virtual block devices, then each block device will need hundreds of threads and possibly also hundreds of connecti

Re: [Qemu-discuss] ESXi prebuilt image

2017-08-06 Thread Jan Schermer
You can’t redistribute that image legally, which is probably why you didn’t find any. Solution: virt-install --connect=qemu:///system -n esxi --cpu=host --vcpus=2 \ --ram 4096 --os-type=linux \ --cdrom=/tank/images/VMware-VMvisor-Installer-6.5.0.update01-5969303.x86_64.iso \ --disk \ path=/dev/

Re: [Qemu-discuss] ESXi prebuilt image

2017-08-07 Thread Jan Schermer
with your My VMware Super User, Procurement Contact or Administrator. > > If you recently purchased this product through VMware Store or through > a third-party, try downloading later." > > Any ideas where it is hosted so me (and others) can download it ? > > Regards, >

Re: [Qemu-discuss] Virtual ccid is empty

2017-09-07 Thread Jan Schermer
Just a wild guess - I played with this shortly a year ago. There are two formats of NSS database and there’s a mismatch between what qemu supports and what my Ubuntu certutil defaults to. I had to set NSS_DEFAULT_DB_TYPE=“sql" (I think?) to make qemu use the new format... or the other way aroun

Re: [Qemu-discuss] Virtual ccid is empty

2017-09-07 Thread Jan Schermer
l' files which are not present in my case. > But they are optional according to my understanding of how sqlite works. > > On 09/07/2017 12:08 PM, Jan Schermer wrote: >> Just a wild guess - I played with this shortly a year ago. There are two >> formats of NSS database a

Re: [Qemu-discuss] Virtual ccid is empty

2017-09-07 Thread Jan Schermer
ccording to strace qemu reads exactly these > files. I've also tried the trick with setting NSS_DEFAULT_DB_TYPE to > 'sql' just in case, didn't help unfortunately. > > Anton > > On 09/07/2017 02:48 PM, Jan Schermer wrote: >> That might be true, but is t

[Qemu-discuss] PCI passthrough error with IGB multiport ethernet

2017-10-25 Thread Jan Schermer
Hi, I’m trying to passthrough two dual-port cards to a guest. It seems to work fine until I actually the cards, then I get: [ 306.667890] pcieport :00:03.0: AER: Multiple Uncorrected (Fatal) error received: id= [ 306.667904] vfio-pci :03:00.0: PCIe Bus Error: severity=Uncorrected (

Re: [Qemu-discuss] Issue related to mounting /dev/nbd0p1

2018-12-04 Thread Jan Schermer
Partigion is missing :-) try: kpartx -a /dev/ndb0 > On 4 Dec 2018, at 11:35, ramakanth varala wrote: > > Hi Pascal.. > > yes its RW ... > > [root@localhost ~]# fdisk -l /dev/nbd0 > > Disk /dev/nbd0: 21.4 GB, 21474836480 bytes > 255 heads, 63 sectors/track, 2610 cylinders > Units = cylinders

[Qemu-discuss] GPU passthrough not working

2019-06-16 Thread Jan Schermer
Hi, hoping someone can help with this. I ran into several snags trying to setup GPU passthrough on my box. It is the only GPU, thus primary. efifb claims mem region on the GPU so it has to be blacklisted completely, even disabling it with video=efifb:off is not enough, the console would no longe