ia64/linux-2.6.18-xen.hg

view Documentation/networking/pt.txt @ 897:329ea0ccb344

balloon: try harder to balloon up under memory pressure.

Currently if the balloon driver is unable to increase the guest's
reservation it assumes the failure was due to reaching its full
allocation, gives up on the ballooning operation and records the limit
it reached as the "hard limit". The driver will not try again until
the target is set again (even to the same value).

However it is possible that ballooning has in fact failed due to
memory pressure in the host and therefore it is desirable to keep
attempting to reach the target in case memory becomes available. The
most likely scenario is that some guests are ballooning down while
others are ballooning up and therefore there is temporary memory
pressure while things stabilise. You would not expect a well behaved
toolstack to ask a domain to balloon to more than its allocation nor
would you expect it to deliberately over-commit memory by setting
balloon targets which exceed the total host memory.

This patch drops the concept of a hard limit and causes the balloon
driver to retry increasing the reservation on a timer in the same
manner as when decreasing the reservation.

Also if we partially succeed in increasing the reservation
(i.e. receive less pages than we asked for) then we may as well keep
those pages rather than returning them to Xen.

Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
author Keir Fraser <keir.fraser@citrix.com>
date Fri Jun 05 14:01:20 2009 +0100 (2009-06-05)
parents 831230e53067
children
line source
1 This is the README for the Gracilis Packetwin device driver, version 0.5
2 ALPHA for Linux 1.3.43.
4 These files will allow you to talk to the PackeTwin (now know as PT) and
5 connect through it just like a pair of TNCs. To do this you will also
6 require the AX.25 code in the kernel enabled.
8 There are four files in this archive; this readme, a patch file, a .c file
9 and finally a .h file. The two program files need to be put into the
10 drivers/net directory in the Linux source tree, for me this is the
11 directory /usr/src/linux/drivers/net. The patch file needs to be patched in
12 at the top of the Linux source tree (/usr/src/linux in my case).
14 You will most probably have to edit the pt.c file to suit your own setup,
15 this should just involve changing some of the defines at the top of the file.
16 Please note that if you run an external modem you must specify a speed of 0.
18 The program is currently setup to run a 4800 baud external modem on port A
19 and a Kantronics DE-9600 daughter board on port B so if you have this (or
20 something similar) then you're right.
22 To compile in the driver, put the files in the correct place and patch in
23 the diff. You will have to re-configure the kernel again before you
24 recompile it.
26 The driver is not real good at the moment for finding the card. You can
27 'help' it by changing the order of the potential addresses in the structure
28 found in the pt_init() function so the address of where the card is is put
29 first.
31 After compiling, you have to get them going, they are pretty well like any
32 other net device and just need ifconfig to get them going.
33 As an example, here is my /etc/rc.net
34 --------------------------
36 #
37 # Configure the PackeTwin, port A.
38 /sbin/ifconfig pt0a 44.136.8.87 hw ax25 vk2xlz mtu 512
39 /sbin/ifconfig pt0a 44.136.8.87 broadcast 44.136.8.255 netmask 255.255.255.0
40 /sbin/route add -net 44.136.8.0 netmask 255.255.255.0 dev pt0a
41 /sbin/route add -net 44.0.0.0 netmask 255.0.0.0 gw 44.136.8.68 dev pt0a
42 /sbin/route add -net 138.25.16.0 netmask 255.255.240.0 dev pt0a
43 /sbin/route add -host 44.136.8.255 dev pt0a
44 #
45 # Configure the PackeTwin, port B.
46 /sbin/ifconfig pt0b 44.136.8.87 hw ax25 vk2xlz-1 mtu 512
47 /sbin/ifconfig pt0b 44.136.8.87 broadcast 44.255.255.255 netmask 255.0.0.0
48 /sbin/route add -host 44.136.8.216 dev pt0b
49 /sbin/route add -host 44.136.8.95 dev pt0b
50 /sbin/route add -host 44.255.255.255 dev pt0b
52 This version of the driver comes under the GNU GPL. If you have one of my
53 previous (non-GPL) versions of the driver, please update to this one.
55 I hope that this all works well for you. I would be pleased to hear how
56 many people use the driver and if it does its job.
58 - Craig vk2xlz <csmall@small.dropbear.id.au>