XL vs Xend Feature Comparison: Difference between revisions

Revision as of 09:55, 9 August 2012

The following compares the featureset of the xend toolstack vs the xl toolstack as of Xen 4.2.

	xm / xend	xl
Guest Types
PV guests	✓	✓
HVM guests	✓	✓
Guest lifecycle operations
Create, shutdown, reboot	✓	✓
Pause, unpause	✓	✓
List, rename	✓	✓
Live migration and save/restore	✓	✓
Guest device support
Network devices (both para-virtualised and emulated)	✓	✓
Block devices (both para-virtualised and emulated)	✓	✓
Virtual framebuffer, keyboard and mouse (both para-virtualised and emulated)	✓ (includes SDL & VNC support)	✓ (includes SDL, VNC & SPICE support)
PV console devices	✗	✓
PCI passthrough	✓	✓
Device Model Support
Qemu Xen Traditional Device Model	✓	✓
Upstream Qemu Device Model	✗	✓
Stub domain Qemu Xen Traditional Device Model	✓	✓
Stub domain Upstream Qemu Device Model	✗	✗

CPU pools	???	✓
Cpupool-per-NUMA node.	✗	✓
Control of CPU scheduler parameters	✓	✓
TMEM	✓	✓
XSM/Flash policy control	✓	✓
Control of CPUID features exposed to guests.	✓	✓ (syntax greatly improved)

Managed Domains	✓	No ^{[ 1 ]}
Python code in configuration files	✓	No ^{[ 1 ]}

^{[ 1 ]} Feature is explicitly not supported by xl, see XL Anti-Features

@@ Line 49: / Line 49: @@
 |-
 |PV console devices
+|{{NotDone}}
-|
 |{{Tick}}
 |-
@@ Line 63: / Line 63: @@
 |-
 |Upstream Qemu Device Model
+|{{NotDone}}
-|
 |{{Tick}}
 |-
@@ Line 71: / Line 71: @@
 |-
 |Stub domain Upstream Qemu Device Model
+|{{NotDone}}
-|
+|{{NotDone}}
-|
 |-
 !colspan="3"|<big>&nbsp;</big>
@@ Line 81: / Line 81: @@
 |-
 |Cpupool-per-NUMA node.
+|{{NotDone}}
-|
 |{{Tick}}
 |-