#xenserver — Public Fediverse posts

Updated to #XenServer 9 today and all my SCSI storage repositories were broken (both iSCSI and FibreChannel). It turns out that if you have a USB memory stick plugged in, it kills multipathing on the whole system.

# multipath -ll
sde: broken device without vendor ID
# multipath -v9
[snip]
596.213238 | sde: vendor = UNKNOWN
596.213281 | sde: product = USB DISK 3.0
596.213323 | sde: rev = PMAP
596.214049 | find_hwe: found 0 hwtable matches for UNKNOWN:USB DISK 3.0:PMAP

The fix? Unplug the USB 🤬

Updated to #XenServer 9 today and all my SCSI storage repositories were broken (both iSCSI and FibreChannel). It turns out that if you have a USB memory stick plugged in, it kills multipathing on the whole system.

# multipath -ll
sde: broken device without vendor ID
# multipath -v9
[snip]
596.213238 | sde: vendor = UNKNOWN
596.213281 | sde: product = USB DISK 3.0
596.213323 | sde: rev = PMAP
596.214049 | find_hwe: found 0 hwtable matches for UNKNOWN:USB DISK 3.0:PMAP

The fix? Unplug the USB 🤬

Updated to #XenServer 9 today and all my SCSI storage repositories were broken (both iSCSI and FibreChannel). It turns out that if you have a USB memory stick plugged in, it kills multipathing on the whole system.

# multipath -ll
sde: broken device without vendor ID
# multipath -v9
[snip]
596.213238 | sde: vendor = UNKNOWN
596.213281 | sde: product = USB DISK 3.0
596.213323 | sde: rev = PMAP
596.214049 | find_hwe: found 0 hwtable matches for UNKNOWN:USB DISK 3.0:PMAP

The fix? Unplug the USB 🤬

Updated to #XenServer 9 today and all my SCSI storage repositories were broken (both iSCSI and FibreChannel). It turns out that if you have a USB memory stick plugged in, it kills multipathing on the whole system.

# multipath -ll
sde: broken device without vendor ID
# multipath -v9
[snip]
596.213238 | sde: vendor = UNKNOWN
596.213281 | sde: product = USB DISK 3.0
596.213323 | sde: rev = PMAP
596.214049 | find_hwe: found 0 hwtable matches for UNKNOWN:USB DISK 3.0:PMAP

The fix? Unplug the USB 🤬

Updated to #XenServer 9 today and all my SCSI storage repositories were broken (both iSCSI and FibreChannel). It turns out that if you have a USB memory stick plugged in, it kills multipathing on the whole system.

# multipath -ll
sde: broken device without vendor ID
# multipath -v9
[snip]
596.213238 | sde: vendor = UNKNOWN
596.213281 | sde: product = USB DISK 3.0
596.213323 | sde: rev = PMAP
596.214049 | find_hwe: found 0 hwtable matches for UNKNOWN:USB DISK 3.0:PMAP

The fix? Unplug the USB 🤬

An interesting article explaining why #Citrix #XenServer 9 won't run in my lab:
https://www.linkedin.com/pulse/part-1-xenserver-9-x86-64-v3-wall-artur-ogloza-e8n8f/

It's down to a hard requirement for x86-64-v3 (I have avx2, bmi2 and fma missing from /proc/cpuinfo).

An interesting article explaining why #Citrix #XenServer 9 won't run in my lab:
https://www.linkedin.com/pulse/part-1-xenserver-9-x86-64-v3-wall-artur-ogloza-e8n8f/

It's down to a hard requirement for x86-64-v3 (I have avx2, bmi2 and fma missing from /proc/cpuinfo).

An interesting article explaining why #Citrix #XenServer 9 won't run in my lab:
https://www.linkedin.com/pulse/part-1-xenserver-9-x86-64-v3-wall-artur-ogloza-e8n8f/

It's down to a hard requirement for x86-64-v3 (I have avx2, bmi2 and fma missing from /proc/cpuinfo).

An interesting article explaining why #Citrix #XenServer 9 won't run in my lab:
https://www.linkedin.com/pulse/part-1-xenserver-9-x86-64-v3-wall-artur-ogloza-e8n8f/

It's down to a hard requirement for x86-64-v3 (I have avx2, bmi2 and fma missing from /proc/cpuinfo).

An interesting article explaining why #Citrix #XenServer 9 won't run in my lab:
https://www.linkedin.com/pulse/part-1-xenserver-9-x86-64-v3-wall-artur-ogloza-e8n8f/

It's down to a hard requirement for x86-64-v3 (I have avx2, bmi2 and fma missing from /proc/cpuinfo).

citrix XenServer 9 has been published today.

be aware of some changes:
* systemd "predictable" network interface names in dom0
* Host BIOS Boot Support has been removed
* Linux Bridge Network Stack has been removed
https://docs.xenserver.com/en-us/xenserver/9/whats-new
#citrix #XenServer #Xen

citrix XenServer 9 has been published today.

be aware of some changes:
* systemd "predictable" network interface names in dom0
* Host BIOS Boot Support has been removed
* Linux Bridge Network Stack has been removed
https://docs.xenserver.com/en-us/xenserver/9/whats-new
#citrix #XenServer #Xen

citrix XenServer 9 has been published today.

be aware of some changes:
* systemd "predictable" network interface names in dom0
* Host BIOS Boot Support has been removed
* Linux Bridge Network Stack has been removed
https://docs.xenserver.com/en-us/xenserver/9/whats-new
#citrix #XenServer #Xen

citrix XenServer 9 has been published today.

be aware of some changes:
* systemd "predictable" network interface names in dom0
* Host BIOS Boot Support has been removed
* Linux Bridge Network Stack has been removed
https://docs.xenserver.com/en-us/xenserver/9/whats-new
#citrix #XenServer #Xen

Yes, it's the same with @xcpng as #XenServer. XenCenter and XCP-ng Center don't offer the SR-IOV network as an option for VMs. Luckily, #XenOrchestra gets that right.

While there I tested various #BSD OSes with SR-IOV and Intel i350 NICs:

#NetBSD - NIC is described (as defined in pcidevs), but no driver
#OpenBSD - NIC entirely unknown as not in pcidevs
#FreeBSD - virtual function works straight off as igb0

The #NetBSD and slightly-newer #OpenBSD iavf(8) drivers only support much newer NICs.

Yes, it's the same with @xcpng as #XenServer. XenCenter and XCP-ng Center don't offer the SR-IOV network as an option for VMs. Luckily, #XenOrchestra gets that right.

While there I tested various #BSD OSes with SR-IOV and Intel i350 NICs:

#NetBSD - NIC is described (as defined in pcidevs), but no driver
#OpenBSD - NIC entirely unknown as not in pcidevs
#FreeBSD - virtual function works straight off as igb0

The #NetBSD and slightly-newer #OpenBSD iavf(8) drivers only support much newer NICs.

Yes, it's the same with @xcpng as #XenServer. XenCenter and XCP-ng Center don't offer the SR-IOV network as an option for VMs. Luckily, #XenOrchestra gets that right.

While there I tested various #BSD OSes with SR-IOV and Intel i350 NICs:

#NetBSD - NIC is described (as defined in pcidevs), but no driver
#OpenBSD - NIC entirely unknown as not in pcidevs
#FreeBSD - virtual function works straight off as igb0

The #NetBSD and slightly-newer #OpenBSD iavf(8) drivers only support much newer NICs.

Yes, it's the same with @xcpng as #XenServer. XenCenter and XCP-ng Center don't offer the SR-IOV network as an option for VMs. Luckily, #XenOrchestra gets that right.

While there I tested various #BSD OSes with SR-IOV and Intel i350 NICs:

#NetBSD - NIC is described (as defined in pcidevs), but no driver
#OpenBSD - NIC entirely unknown as not in pcidevs
#FreeBSD - virtual function works straight off as igb0

The #NetBSD and slightly-newer #OpenBSD iavf(8) drivers only support much newer NICs.

Yes, it's the same with @xcpng as #XenServer. XenCenter and XCP-ng Center don't offer the SR-IOV network as an option for VMs. Luckily, #XenOrchestra gets that right.

While there I tested various #BSD OSes with SR-IOV and Intel i350 NICs:

#NetBSD - NIC is described (as defined in pcidevs), but no driver
#OpenBSD - NIC entirely unknown as not in pcidevs
#FreeBSD - virtual function works straight off as igb0

The #NetBSD and slightly-newer #OpenBSD iavf(8) drivers only support much newer NICs.

Well, that was annoying. I enabled SR-IOV in #XenServer and it stopped traffic to all VLANs on all interfaces. This is fair enough on the physical interface that SR-IOV is enabled on, but I wasn't expecting it to affect all NICs.

I wasted a lot of time hunting through switch and router configs to track this down.

Disabling SR-IOV isn't sufficient, each host needs to be rebooted afterwards. Let me see if I can recreate with @xcpng

Well, that was annoying. I enabled SR-IOV in #XenServer and it stopped traffic to all VLANs on all interfaces. This is fair enough on the physical interface that SR-IOV is enabled on, but I wasn't expecting it to affect all NICs.

I wasted a lot of time hunting through switch and router configs to track this down.

Disabling SR-IOV isn't sufficient, each host needs to be rebooted afterwards. Let me see if I can recreate with @xcpng

Well, that was annoying. I enabled SR-IOV in #XenServer and it stopped traffic to all VLANs on all interfaces. This is fair enough on the physical interface that SR-IOV is enabled on, but I wasn't expecting it to affect all NICs.

I wasted a lot of time hunting through switch and router configs to track this down.

Disabling SR-IOV isn't sufficient, each host needs to be rebooted afterwards. Let me see if I can recreate with @xcpng

Well, that was annoying. I enabled SR-IOV in #XenServer and it stopped traffic to all VLANs on all interfaces. This is fair enough on the physical interface that SR-IOV is enabled on, but I wasn't expecting it to affect all NICs.

I wasted a lot of time hunting through switch and router configs to track this down.

Disabling SR-IOV isn't sufficient, each host needs to be rebooted afterwards. Let me see if I can recreate with @xcpng

Well, that was annoying. I enabled SR-IOV in #XenServer and it stopped traffic to all VLANs on all interfaces. This is fair enough on the physical interface that SR-IOV is enabled on, but I wasn't expecting it to affect all NICs.

I wasted a lot of time hunting through switch and router configs to track this down.

Disabling SR-IOV isn't sufficient, each host needs to be rebooted afterwards. Let me see if I can recreate with @xcpng

Busy week upgrading all VMs to #NetBSD 10.1 to allow switching to HVM from PV (as #XenServer and @xcpng have dropped support for PV and unfortunately do not support PVH either) which means moving to uEFI booting rather than pygrub loading the kernel. Just waiting for the next #pkgsrc quarterly release to move the remainder over with nice fresh packages.

Busy week upgrading all VMs to #NetBSD 10.1 to allow switching to HVM from PV (as #XenServer and @xcpng have dropped support for PV and unfortunately do not support PVH either) which means moving to uEFI booting rather than pygrub loading the kernel. Just waiting for the next #pkgsrc quarterly release to move the remainder over with nice fresh packages.

Busy week upgrading all VMs to #NetBSD 10.1 to allow switching to HVM from PV (as #XenServer and @xcpng have dropped support for PV and unfortunately do not support PVH either) which means moving to uEFI booting rather than pygrub loading the kernel. Just waiting for the next #pkgsrc quarterly release to move the remainder over with nice fresh packages.

Busy week upgrading all VMs to #NetBSD 10.1 to allow switching to HVM from PV (as #XenServer and @xcpng have dropped support for PV and unfortunately do not support PVH either) which means moving to uEFI booting rather than pygrub loading the kernel. Just waiting for the next #pkgsrc quarterly release to move the remainder over with nice fresh packages.

Busy week upgrading all VMs to #NetBSD 10.1 to allow switching to HVM from PV (as #XenServer and @xcpng have dropped support for PV and unfortunately do not support PVH either) which means moving to uEFI booting rather than pygrub loading the kernel. Just waiting for the next #pkgsrc quarterly release to move the remainder over with nice fresh packages.

XenServer VM Tools vulnerabilities let attackers gain kernel privileges on Windows VMs. If you run virtualized Windows environments, patch NOW!

#SecurityLand #CyberWatch #XenServer #CVE #VirtualizationSecurity #Windows #Vulnerability

https://www.security.land/three-lines-of-code-that-could-compromise-your-entire-virtual-infrastructure/

@awinkler Übrigens: OpenMediaVault läuft bei mir in einer #XenServer VM.

Man sieht auch schön die grafische Benutzerführung für Updates & Co.

@awinkler Übrigens: OpenMediaVault läuft bei mir in einer #XenServer VM.

Man sieht auch schön die grafische Benutzerführung für Updates & Co.

@awinkler Übrigens: OpenMediaVault läuft bei mir in einer #XenServer VM.

Man sieht auch schön die grafische Benutzerführung für Updates & Co.

@awinkler Übrigens: OpenMediaVault läuft bei mir in einer #XenServer VM.

Man sieht auch schön die grafische Benutzerführung für Updates & Co.

@awinkler Übrigens: OpenMediaVault läuft bei mir in einer #XenServer VM.

Man sieht auch schön die grafische Benutzerführung für Updates & Co.

Today I've been playing with #pkgsrc on #FreeBSD and trying to get some of my #NetBSD packages working on it.

pkgsrc bootstrap worked well on FreeBSD. Hardest part is remember you can't just type make any more!

I've been focussing on the #XenServer tools which looks to be rather stagnant in FreeBSD ports. The #pkgsrc version will work better on FreeBSD than the version from ports.

Just need to work out what to do about rc.d scripts. I can't find any prior art for rc.d from pkgsrc on FreeBSD

Today I've been playing with #pkgsrc on #FreeBSD and trying to get some of my #NetBSD packages working on it.

pkgsrc bootstrap worked well on FreeBSD. Hardest part is remember you can't just type make any more!

I've been focussing on the #XenServer tools which looks to be rather stagnant in FreeBSD ports. The #pkgsrc version will work better on FreeBSD than the version from ports.

Just need to work out what to do about rc.d scripts. I can't find any prior art for rc.d from pkgsrc on FreeBSD

Today I've been playing with #pkgsrc on #FreeBSD and trying to get some of my #NetBSD packages working on it.

pkgsrc bootstrap worked well on FreeBSD. Hardest part is remember you can't just type make any more!

I've been focussing on the #XenServer tools which looks to be rather stagnant in FreeBSD ports. The #pkgsrc version will work better on FreeBSD than the version from ports.

Just need to work out what to do about rc.d scripts. I can't find any prior art for rc.d from pkgsrc on FreeBSD

Today I've been playing with #pkgsrc on #FreeBSD and trying to get some of my #NetBSD packages working on it.

pkgsrc bootstrap worked well on FreeBSD. Hardest part is remember you can't just type make any more!

I've been focussing on the #XenServer tools which looks to be rather stagnant in FreeBSD ports. The #pkgsrc version will work better on FreeBSD than the version from ports.

Just need to work out what to do about rc.d scripts. I can't find any prior art for rc.d from pkgsrc on FreeBSD

Today I've been playing with #pkgsrc on #FreeBSD and trying to get some of my #NetBSD packages working on it.

pkgsrc bootstrap worked well on FreeBSD. Hardest part is remember you can't just type make any more!

I've been focussing on the #XenServer tools which looks to be rather stagnant in FreeBSD ports. The #pkgsrc version will work better on FreeBSD than the version from ports.

Just need to work out what to do about rc.d scripts. I can't find any prior art for rc.d from pkgsrc on FreeBSD

Cloud Software Group, owner of Citrix, Tibco, NetScaler, and XenServer, confirms global layoffs, with LinkedIn posts revealing cuts to engineers, technical account managers, and other roles. #CloudSoftwareGroup #Citrix #Tibco #NetScaler #XenServer #Layoffs #TechNews #JobCuts #WorkforceChanges

Cloud Software Group, owner of Citrix, Tibco, NetScaler, and XenServer, confirms global layoffs, with LinkedIn posts revealing cuts to engineers, technical account managers, and other roles. #CloudSoftwareGroup #Citrix #Tibco #NetScaler #XenServer #Layoffs #TechNews #JobCuts #WorkforceChanges

Cloud Software Group, owner of Citrix, Tibco, NetScaler, and XenServer, confirms global layoffs, with LinkedIn posts revealing cuts to engineers, technical account managers, and other roles. #CloudSoftwareGroup #Citrix #Tibco #NetScaler #XenServer #Layoffs #TechNews #JobCuts #WorkforceChanges

Cloud Software Group, owner of Citrix, Tibco, NetScaler, and XenServer, confirms global layoffs, with LinkedIn posts revealing cuts to engineers, technical account managers, and other roles. #CloudSoftwareGroup #Citrix #Tibco #NetScaler #XenServer #Layoffs #TechNews #JobCuts #WorkforceChanges

Cloud Software Group, owner of Citrix, Tibco, NetScaler, and XenServer, confirms global layoffs, with LinkedIn posts revealing cuts to engineers, technical account managers, and other roles. #CloudSoftwareGroup #Citrix #Tibco #NetScaler #XenServer #Layoffs #TechNews #JobCuts #WorkforceChanges

Pretty pleased with the speed of #zfs replication between two #NetBSD systems between our two datacentres in Cambridge:

receiving full stream of tank/[email protected]:01:01 into tank/backup/[email protected]:01:01
received 325GB stream in 3453 seconds (96.4MB/sec)

One end is virtual (a #Xen domU running on #XenServer on a host connected by 16Gb FC to an IBM FlashSystem) and the other is physical (HP Microserver Gen 10). Running over rsh on a private VLAN (so no ssh encryption overhead)

Pretty pleased with the speed of #zfs replication between two #NetBSD systems between our two datacentres in Cambridge:

receiving full stream of tank/[email protected]:01:01 into tank/backup/[email protected]:01:01
received 325GB stream in 3453 seconds (96.4MB/sec)

One end is virtual (a #Xen domU running on #XenServer on a host connected by 16Gb FC to an IBM FlashSystem) and the other is physical (HP Microserver Gen 10). Running over rsh on a private VLAN (so no ssh encryption overhead)

Pretty pleased with the speed of #zfs replication between two #NetBSD systems between our two datacentres in Cambridge:

receiving full stream of tank/[email protected]:01:01 into tank/backup/[email protected]:01:01
received 325GB stream in 3453 seconds (96.4MB/sec)

One end is virtual (a #Xen domU running on #XenServer on a host connected by 16Gb FC to an IBM FlashSystem) and the other is physical (HP Microserver Gen 10). Running over rsh on a private VLAN (so no ssh encryption overhead)

Pretty pleased with the speed of #zfs replication between two #NetBSD systems between our two datacentres in Cambridge:

receiving full stream of tank/[email protected]:01:01 into tank/backup/[email protected]:01:01
received 325GB stream in 3453 seconds (96.4MB/sec)

One end is virtual (a #Xen domU running on #XenServer on a host connected by 16Gb FC to an IBM FlashSystem) and the other is physical (HP Microserver Gen 10). Running over rsh on a private VLAN (so no ssh encryption overhead)

Pretty pleased with the speed of #zfs replication between two #NetBSD systems between our two datacentres in Cambridge:

receiving full stream of tank/[email protected]:01:01 into tank/backup/[email protected]:01:01
received 325GB stream in 3453 seconds (96.4MB/sec)

One end is virtual (a #Xen domU running on #XenServer on a host connected by 16Gb FC to an IBM FlashSystem) and the other is physical (HP Microserver Gen 10). Running over rsh on a private VLAN (so no ssh encryption overhead)

@mWare @ltning Many of my #ZFS installs are virtual, maybe even on #XenServer connected to an IBM #FlashSystem over FibreChannel. There's do many levels of abstraction, no option to offload a ZIL

@mWare @ltning Many of my #ZFS installs are virtual, maybe even on #XenServer connected to an IBM #FlashSystem over FibreChannel. There's do many levels of abstraction, no option to offload a ZIL

@mWare @ltning Many of my #ZFS installs are virtual, maybe even on #XenServer connected to an IBM #FlashSystem over FibreChannel. There's do many levels of abstraction, no option to offload a ZIL

@mWare @ltning Many of my #ZFS installs are virtual, maybe even on #XenServer connected to an IBM #FlashSystem over FibreChannel. There's do many levels of abstraction, no option to offload a ZIL

@mWare @ltning Many of my #ZFS installs are virtual, maybe even on #XenServer connected to an IBM #FlashSystem over FibreChannel. There's do many levels of abstraction, no option to offload a ZIL

@willnix Try #XCPng. It is a fork of the old #XenServer code before it went closed source. I noticed you had #Citrix in your tags, so it could be a good fit for you.

Can I do a shout-out to my favourite #hypervisor #xcp-ng (https://xcp-ng.org/)? Fast install times, faster machines, easy to manage, disaster recovers is a bliss. And it is rock-solid.

I have been using it (an #xenserver initially) for more than a decade.

#opensource #virtualisation #vms #linux #server

Can I do a shout-out to my favourite #hypervisor #xcp-ng (https://xcp-ng.org/)? Fast install times, faster machines, easy to manage, disaster recovers is a bliss. And it is rock-solid.

I have been using it (an #xenserver initially) for more than a decade.

#opensource #virtualisation #vms #linux #server

Can I do a shout-out to my favourite #hypervisor #xcp-ng (https://xcp-ng.org/)? Fast install times, faster machines, easy to manage, disaster recovers is a bliss. And it is rock-solid.

I have been using it (an #xenserver initially) for more than a decade.

#opensource #virtualisation #vms #linux #server

Can I do a shout-out to my favourite #hypervisor #xcp-ng (https://xcp-ng.org/)? Fast install times, faster machines, easy to manage, disaster recovers is a bliss. And it is rock-solid.

I have been using it (an #xenserver initially) for more than a decade.

#opensource #virtualisation #vms #linux #server