Peter Van Biesen
2008-Oct-07 08:27 UTC
[Xen-users] Windows XP filesystem corruption when using device-mapper ?
Hi, I''m currently running windows XP without the gplpv drivers. At nearly every reboot, some directory or another gets corrupted on the filesystem. I found bug 1363, which talks about gplpv and lvm, but maybe gplpv is not the problem here, maybe it is the ioemu that is at fault. Anyway, I had a lot of problems with gplpv on lvm. When I used the plain disk, windows would boot and run, but sometimes it would want to do a disk check on reboot. Letting it do that would corrupt the fs again. Beware: when I talk about ''plain disk'', I mean a device mapper device created by the multipath daemon. So the dm module is still in between it. I Haven''t tried using a real plain disk ( a disk actually in the server ). The dom0 is running on a 2.6.18, maybe it is a bug in the devicemapper of this old version ? Does anybody have the same experience using multipath/lvm/device mapper ? Kindest regards, Peter. -- Peter Van Biesen Sysadmin VAPH tel: +32 (0) 2 225 85 70 fax: +32 (0) 2 225 85 88 e-mail: peter.vanbiesen@vaph.be PGP: http://www.vaph.be/pgpkeys _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
Nemeth, Tamas
2008-Oct-07 09:52 UTC
Re: [Xen-users] Windows XP filesystem corruption when using device-mapper ?
Hi, how are you doing? 2008. 10. 7, kedd keltezéssel 10.27-kor Peter Van Biesen ezt írta:> I''m currently running windows XP without the gplpv drivers.Have you installed those drivers? I so, have you ever booted with the /gplpv option? What is your domU config? Best wishes, Tamas _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
James Harper
2008-Oct-07 10:39 UTC
RE: [Xen-users] Windows XP filesystem corruption when usingdevice-mapper ?
> Hi, how are you doing? > > 2008. 10. 7, kedd keltezéssel 10.27-kor Peter Van Biesen ezt írta: > > I''m currently running windows XP without the gplpv drivers. > > Have you installed those drivers? I so, have you ever booted with > the /gplpv option? What is your domU config? >If his system is crashing without the drivers then I have doubts that installing them will make things any better. The only thing they might do is that they might be less likely to cause corruption if the physical server crashes, and I haven''t done enough testing to say that with any certainty. James _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users
Peter Van Biesen
2008-Oct-08 10:45 UTC
Re: [Xen-users] Windows XP filesystem corruption when usingdevice-mapper ?
Hi,
the windows xp doesn''t crash. It runs fine. But rebooting is always
kind of a gamble - what will be missing ? boot.ini ? ntoskrnl.exe ? or does
windows xp just want to chkdsk ? I''ve tried the gplpv drivers, they
seem to run but switching between gplpv and non-gplpv would break windows
completely. If I try to change networksettings of the xen network card while in
gplpv mode windows systematically breaks.
I threw away that windows xp image. I used it to try live migration and this
does not work - gplpv or not ( I''ve given up on that ). I now have a
freshly installed windows xp image running without gplpv and a second image
running with gplpv drivers.
Without the gplpv drivers, it now runs smoothly but slow for disk io. With the
drivers, my rdp session stops working every couple of minutes. When I run a ping
on it, the connection is very very intermittent and sometime stalls or drops
away. I use dhcp for both the ioemu and the xen network adapter, so it should be
identical.
64 bytes from 172.16.6.231: icmp_seq=14 ttl=128 time=0.242 ms
64 bytes from 172.16.6.231: icmp_seq=15 ttl=128 time=0.252 ms
64 bytes from 172.16.6.231: icmp_seq=16 ttl=128 time=0.273 ms
64 bytes from 172.16.6.231: icmp_seq=17 ttl=128 time=0.262 ms
64 bytes from 172.16.6.231: icmp_seq=18 ttl=128 time=0.238 ms
64 bytes from 172.16.6.231: icmp_seq=19 ttl=128 time=0.261 ms
64 bytes from 172.16.6.231: icmp_seq=20 ttl=128 time=0.241 ms
64 bytes from 172.16.6.231: icmp_seq=21 ttl=128 time=0.247 ms
64 bytes from 172.16.6.231: icmp_seq=22 ttl=128 time=0.238 ms
64 bytes from 172.16.6.231: icmp_seq=23 ttl=128 time=5887 ms
64 bytes from 172.16.6.231: icmp_seq=24 ttl=128 time=4887 ms
64 bytes from 172.16.6.231: icmp_seq=25 ttl=128 time=3887 ms
64 bytes from 172.16.6.231: icmp_seq=26 ttl=128 time=2887 ms
64 bytes from 172.16.6.231: icmp_seq=27 ttl=128 time=1887 ms
64 bytes from 172.16.6.231: icmp_seq=28 ttl=128 time=887 ms
64 bytes from 172.16.6.231: icmp_seq=29 ttl=128 time=0.281 ms
64 bytes from 172.16.6.231: icmp_seq=30 ttl=128 time=0.252 ms
64 bytes from 172.16.6.231: icmp_seq=31 ttl=128 time=0.266 ms
64 bytes from 172.16.6.231: icmp_seq=32 ttl=128 time=0.261 ms
64 bytes from 172.16.6.231: icmp_seq=33 ttl=128 time=0.273 ms
64 bytes from 172.16.6.231: icmp_seq=34 ttl=128 time=0.258 ms
64 bytes from 172.16.6.231: icmp_seq=35 ttl=128 time=0.255 ms
64 bytes from 172.16.6.231: icmp_seq=36 ttl=128 time=0.247 ms
64 bytes from 172.16.6.231: icmp_seq=37 ttl=128 time=0.263 ms
From pcvlf1922.local (172.16.4.70) icmp_seq=129 Destination Host Unreachable
From pcvlf1922.local (172.16.4.70) icmp_seq=130 Destination Host Unreachable
From pcvlf1922.local (172.16.4.70) icmp_seq=131 Destination Host Unreachable
From pcvlf1922.local (172.16.4.70) icmp_seq=132 Destination Host Unreachable
64 bytes from 172.16.6.231: icmp_seq=58 ttl=128 time=77914 ms
From pcvlf1922.local (172.16.4.70) icmp_seq=133 Destination Host Unreachable
From pcvlf1922.local (172.16.4.70) icmp_seq=134 Destination Host Unreachable
From pcvlf1922.local (172.16.4.70) icmp_seq=135 Destination Host Unreachable
64 bytes from 172.16.6.231: icmp_seq=59 ttl=128 time=76914 ms
64 bytes from 172.16.6.231: icmp_seq=60 ttl=128 time=75914 ms
64 bytes from 172.16.6.231: icmp_seq=61 ttl=128 time=74914 ms
64 bytes from 172.16.6.231: icmp_seq=62 ttl=128 time=73914 ms
64 bytes from 172.16.6.231: icmp_seq=63 ttl=128 time=72914 ms
64 bytes from 172.16.6.231: icmp_seq=64 ttl=128 time=71914 ms
64 bytes from 172.16.6.231: icmp_seq=65 ttl=128 time=70914 ms
64 bytes from 172.16.6.231: icmp_seq=66 ttl=128 time=69914 ms
I also am running linux domu''s on the same server, they do not have any
problem whatsoever.
My impression is that xen and windows is still highly experimental and
I''m running out of time to prove ( to my bosses ) that it can be done.
I still have to find someone who is running a windows under xen in production
...
We''re running 20 dom0''s and about 50 linux domU''s in
production at the moment without problems but if one would ask me how to
virtualise a windows, I would have to advise vmware. From my ( sysadmin ) point
of view there are to many ''strange things'' happening.
Something strange i saw that differs from the linux''es is the tap*
devices. They tend to change the mac address of the bridge on which they are
attached.
vlanbr640 Link encap:Ethernet HWaddr 1a:2d:54:ad:2e:7e
inet6 addr: fe80::200:ff:fe00:0/64 Scope:Link
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:1845367 errors:0 dropped:0 overruns:0 frame:0
TX packets:3 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:0
RX bytes:119469265 (113.9 MiB) TX bytes:258 (258.0 B)
On the other bridges, the HWaddr is fe:ff:ff:ff:ff:ff .
Btw, I just rebooted my image with gplpv driver, screenshots in attachment.
Kindest regards,
Peter.
On Tuesday 07 October 2008 12:39:10 James Harper wrote:> > Hi, how are you doing?
> >
> > 2008. 10. 7, kedd keltezéssel 10.27-kor Peter Van Biesen ezt írta:
> > > I''m currently running windows XP without the gplpv
drivers.
> >
> > Have you installed those drivers? I so, have you ever booted with
> > the /gplpv option? What is your domU config?
> >
>
> If his system is crashing without the drivers then I have doubts that
installing them will make things any better. The only thing they might do is
that they might be less likely to cause corruption if the physical server
crashes, and I haven''t done enough testing to say that with any
certainty.
>
> James
>
>
--
Peter Van Biesen
Sysadmin VAPH
tel: +32 (0) 2 225 85 70
fax: +32 (0) 2 225 85 88
e-mail: peter.vanbiesen@vaph.be
PGP: http://www.vaph.be/pgpkeys
_______________________________________________
Xen-users mailing list
Xen-users@lists.xensource.com
http://lists.xensource.com/xen-users
Peter Van Biesen
2008-Oct-08 14:06 UTC
Re: [Xen-users] Windows XP filesystem corruption when usingdevice-mapper ?
Sorry, my config: kernel = "/usr/lib/xen-default/boot/hvmloader" builder=''hvm'' memory = 512 shadow_memory = 16 name = "vbxl204" disk = [ ''phy:/dev/disk/by-id/scsi-350060e800351af00000051af0000000d,hda,w'', ''file:/bigvol/installmedia/winxp-en-setup-SP2.iso,hdc:cdrom,r'',''file:/bigvol/installmedia/winxp-SP1a-SP3-dotnet2.0.iso,hdd:cdrom,r'', ] vif = [ ''mac=00:16:3E:00:02:03,bridge=vlanbr640'', ] device_model = ''/usr/lib/xen-default/bin/qemu-dm'' boot=''dc'' sdl=0 vnc=1 vnclisten="0.0.0.0" vncunused=1 acpi=1 apic=0 pae=1 usb = 1 usbdevice = ''tablet'' #serial="pty" Kindest regards, Peter. On Wednesday 08 October 2008 12:45:35 Peter Van Biesen wrote:> Hi, > > the windows xp doesn''t crash. It runs fine. But rebooting is always kind of a gamble - what will be missing ? boot.ini ? ntoskrnl.exe ? or does windows xp just want to chkdsk ? I''ve tried the gplpv drivers, they seem to run but switching between gplpv and non-gplpv would break windows completely. If I try to change networksettings of the xen network card while in gplpv mode windows systematically breaks. > > I threw away that windows xp image. I used it to try live migration and this does not work - gplpv or not ( I''ve given up on that ). I now have a freshly installed windows xp image running without gplpv and a second image running with gplpv drivers. > > Without the gplpv drivers, it now runs smoothly but slow for disk io. With the drivers, my rdp session stops working every couple of minutes. When I run a ping on it, the connection is very very intermittent and sometime stalls or drops away. I use dhcp for both the ioemu and the xen network adapter, so it should be identical. > > 64 bytes from 172.16.6.231: icmp_seq=14 ttl=128 time=0.242 ms > 64 bytes from 172.16.6.231: icmp_seq=15 ttl=128 time=0.252 ms > 64 bytes from 172.16.6.231: icmp_seq=16 ttl=128 time=0.273 ms > 64 bytes from 172.16.6.231: icmp_seq=17 ttl=128 time=0.262 ms > 64 bytes from 172.16.6.231: icmp_seq=18 ttl=128 time=0.238 ms > 64 bytes from 172.16.6.231: icmp_seq=19 ttl=128 time=0.261 ms > 64 bytes from 172.16.6.231: icmp_seq=20 ttl=128 time=0.241 ms > 64 bytes from 172.16.6.231: icmp_seq=21 ttl=128 time=0.247 ms > 64 bytes from 172.16.6.231: icmp_seq=22 ttl=128 time=0.238 ms > 64 bytes from 172.16.6.231: icmp_seq=23 ttl=128 time=5887 ms > 64 bytes from 172.16.6.231: icmp_seq=24 ttl=128 time=4887 ms > 64 bytes from 172.16.6.231: icmp_seq=25 ttl=128 time=3887 ms > 64 bytes from 172.16.6.231: icmp_seq=26 ttl=128 time=2887 ms > 64 bytes from 172.16.6.231: icmp_seq=27 ttl=128 time=1887 ms > 64 bytes from 172.16.6.231: icmp_seq=28 ttl=128 time=887 ms > 64 bytes from 172.16.6.231: icmp_seq=29 ttl=128 time=0.281 ms > 64 bytes from 172.16.6.231: icmp_seq=30 ttl=128 time=0.252 ms > 64 bytes from 172.16.6.231: icmp_seq=31 ttl=128 time=0.266 ms > 64 bytes from 172.16.6.231: icmp_seq=32 ttl=128 time=0.261 ms > 64 bytes from 172.16.6.231: icmp_seq=33 ttl=128 time=0.273 ms > 64 bytes from 172.16.6.231: icmp_seq=34 ttl=128 time=0.258 ms > 64 bytes from 172.16.6.231: icmp_seq=35 ttl=128 time=0.255 ms > 64 bytes from 172.16.6.231: icmp_seq=36 ttl=128 time=0.247 ms > 64 bytes from 172.16.6.231: icmp_seq=37 ttl=128 time=0.263 ms > > From pcvlf1922.local (172.16.4.70) icmp_seq=129 Destination Host Unreachable > From pcvlf1922.local (172.16.4.70) icmp_seq=130 Destination Host Unreachable > From pcvlf1922.local (172.16.4.70) icmp_seq=131 Destination Host Unreachable > From pcvlf1922.local (172.16.4.70) icmp_seq=132 Destination Host Unreachable > 64 bytes from 172.16.6.231: icmp_seq=58 ttl=128 time=77914 ms > From pcvlf1922.local (172.16.4.70) icmp_seq=133 Destination Host Unreachable > From pcvlf1922.local (172.16.4.70) icmp_seq=134 Destination Host Unreachable > From pcvlf1922.local (172.16.4.70) icmp_seq=135 Destination Host Unreachable > 64 bytes from 172.16.6.231: icmp_seq=59 ttl=128 time=76914 ms > 64 bytes from 172.16.6.231: icmp_seq=60 ttl=128 time=75914 ms > 64 bytes from 172.16.6.231: icmp_seq=61 ttl=128 time=74914 ms > 64 bytes from 172.16.6.231: icmp_seq=62 ttl=128 time=73914 ms > 64 bytes from 172.16.6.231: icmp_seq=63 ttl=128 time=72914 ms > 64 bytes from 172.16.6.231: icmp_seq=64 ttl=128 time=71914 ms > 64 bytes from 172.16.6.231: icmp_seq=65 ttl=128 time=70914 ms > 64 bytes from 172.16.6.231: icmp_seq=66 ttl=128 time=69914 ms > > I also am running linux domu''s on the same server, they do not have any problem whatsoever. > > My impression is that xen and windows is still highly experimental and I''m running out of time to prove ( to my bosses ) that it can be done. I still have to find someone who is running a windows under xen in production ... > > We''re running 20 dom0''s and about 50 linux domU''s in production at the moment without problems but if one would ask me how to virtualise a windows, I would have to advise vmware. From my ( sysadmin ) point of view there are to many ''strange things'' happening. > > Something strange i saw that differs from the linux''es is the tap* devices. They tend to change the mac address of the bridge on which they are attached. > > vlanbr640 Link encap:Ethernet HWaddr 1a:2d:54:ad:2e:7e > inet6 addr: fe80::200:ff:fe00:0/64 Scope:Link > UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 > RX packets:1845367 errors:0 dropped:0 overruns:0 frame:0 > TX packets:3 errors:0 dropped:0 overruns:0 carrier:0 > collisions:0 txqueuelen:0 > RX bytes:119469265 (113.9 MiB) TX bytes:258 (258.0 B) > > On the other bridges, the HWaddr is fe:ff:ff:ff:ff:ff . > > Btw, I just rebooted my image with gplpv driver, screenshots in attachment. > > Kindest regards, > > Peter. > On Tuesday 07 October 2008 12:39:10 James Harper wrote: > > > Hi, how are you doing? > > > > > > 2008. 10. 7, kedd keltezéssel 10.27-kor Peter Van Biesen ezt írta: > > > > I''m currently running windows XP without the gplpv drivers. > > > > > > Have you installed those drivers? I so, have you ever booted with > > > the /gplpv option? What is your domU config? > > > > > > > If his system is crashing without the drivers then I have doubts that installing them will make things any better. The only thing they might do is that they might be less likely to cause corruption if the physical server crashes, and I haven''t done enough testing to say that with any certainty. > > > > James > > > > > > >-- Peter Van Biesen Sysadmin VAPH tel: +32 (0) 2 225 85 70 fax: +32 (0) 2 225 85 88 e-mail: peter.vanbiesen@vaph.be PGP: http://www.vaph.be/pgpkeys _______________________________________________ Xen-users mailing list Xen-users@lists.xensource.com http://lists.xensource.com/xen-users