ia64/linux-2.6.18-xen.hg

view Documentation/networking/pktgen.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
3 HOWTO for the linux packet generator
4 ------------------------------------
6 Date: 041221
8 Enable CONFIG_NET_PKTGEN to compile and build pktgen.o either in kernel
9 or as module. Module is preferred. insmod pktgen if needed. Once running
10 pktgen creates a thread on each CPU where each thread has affinty it's CPU.
11 Monitoring and controlling is done via /proc. Easiest to select a suitable
12 a sample script and configure.
14 On a dual CPU:
16 ps aux | grep pkt
17 root 129 0.3 0.0 0 0 ? SW 2003 523:20 [pktgen/0]
18 root 130 0.3 0.0 0 0 ? SW 2003 509:50 [pktgen/1]
21 For montoring and control pktgen creates:
22 /proc/net/pktgen/pgctrl
23 /proc/net/pktgen/kpktgend_X
24 /proc/net/pktgen/ethX
27 Viewing threads
28 ===============
29 /proc/net/pktgen/kpktgend_0
30 Name: kpktgend_0 max_before_softirq: 10000
31 Running:
32 Stopped: eth1
33 Result: OK: max_before_softirq=10000
35 Most important the devices assigend to thread. Note! A device can only belong
36 to one thread.
39 Viewing devices
40 ===============
42 Parm section holds configured info. Current hold running stats.
43 Result is printed after run or after interruption. Example:
45 /proc/net/pktgen/eth1
47 Params: count 10000000 min_pkt_size: 60 max_pkt_size: 60
48 frags: 0 delay: 0 clone_skb: 1000000 ifname: eth1
49 flows: 0 flowlen: 0
50 dst_min: 10.10.11.2 dst_max:
51 src_min: src_max:
52 src_mac: 00:00:00:00:00:00 dst_mac: 00:04:23:AC:FD:82
53 udp_src_min: 9 udp_src_max: 9 udp_dst_min: 9 udp_dst_max: 9
54 src_mac_count: 0 dst_mac_count: 0
55 Flags:
56 Current:
57 pkts-sofar: 10000000 errors: 39664
58 started: 1103053986245187us stopped: 1103053999346329us idle: 880401us
59 seq_num: 10000011 cur_dst_mac_offset: 0 cur_src_mac_offset: 0
60 cur_saddr: 0x10a0a0a cur_daddr: 0x20b0a0a
61 cur_udp_dst: 9 cur_udp_src: 9
62 flows: 0
63 Result: OK: 13101142(c12220741+d880401) usec, 10000000 (60byte,0frags)
64 763292pps 390Mb/sec (390805504bps) errors: 39664
66 Confguring threads and devices
67 ==============================
68 This is done via the /proc interface easiest done via pgset in the scripts
70 Examples:
72 pgset "clone_skb 1" sets the number of copies of the same packet
73 pgset "clone_skb 0" use single SKB for all transmits
74 pgset "pkt_size 9014" sets packet size to 9014
75 pgset "frags 5" packet will consist of 5 fragments
76 pgset "count 200000" sets number of packets to send, set to zero
77 for continuous sends until explicitly stopped.
79 pgset "delay 5000" adds delay to hard_start_xmit(). nanoseconds
81 pgset "dst 10.0.0.1" sets IP destination address
82 (BEWARE! This generator is very aggressive!)
84 pgset "dst_min 10.0.0.1" Same as dst
85 pgset "dst_max 10.0.0.254" Set the maximum destination IP.
86 pgset "src_min 10.0.0.1" Set the minimum (or only) source IP.
87 pgset "src_max 10.0.0.254" Set the maximum source IP.
88 pgset "dst6 fec0::1" IPV6 destination address
89 pgset "src6 fec0::2" IPV6 source address
90 pgset "dstmac 00:00:00:00:00:00" sets MAC destination address
91 pgset "srcmac 00:00:00:00:00:00" sets MAC source address
93 pgset "src_mac_count 1" Sets the number of MACs we'll range through.
94 The 'minimum' MAC is what you set with srcmac.
96 pgset "dst_mac_count 1" Sets the number of MACs we'll range through.
97 The 'minimum' MAC is what you set with dstmac.
99 pgset "flag [name]" Set a flag to determine behaviour. Current flags
100 are: IPSRC_RND #IP Source is random (between min/max),
101 IPDST_RND, UDPSRC_RND,
102 UDPDST_RND, MACSRC_RND, MACDST_RND
104 pgset "udp_src_min 9" set UDP source port min, If < udp_src_max, then
105 cycle through the port range.
107 pgset "udp_src_max 9" set UDP source port max.
108 pgset "udp_dst_min 9" set UDP destination port min, If < udp_dst_max, then
109 cycle through the port range.
110 pgset "udp_dst_max 9" set UDP destination port max.
112 pgset "mpls 0001000a,0002000a,0000000a" set MPLS labels (in this example
113 outer label=16,middle label=32,
114 inner label=0 (IPv4 NULL)) Note that
115 there must be no spaces between the
116 arguments. Leading zeros are required.
117 Do not set the bottom of stack bit,
118 thats done automatically. If you do
119 set the bottom of stack bit, that
120 indicates that you want to randomly
121 generate that address and the flag
122 MPLS_RND will be turned on. You
123 can have any mix of random and fixed
124 labels in the label stack.
126 pgset "mpls 0" turn off mpls (or any invalid argument works too!)
128 pgset stop aborts injection. Also, ^C aborts generator.
131 Example scripts
132 ===============
134 A collection of small tutorial scripts for pktgen is in expamples dir.
136 pktgen.conf-1-1 # 1 CPU 1 dev
137 pktgen.conf-1-2 # 1 CPU 2 dev
138 pktgen.conf-2-1 # 2 CPU's 1 dev
139 pktgen.conf-2-2 # 2 CPU's 2 dev
140 pktgen.conf-1-1-rdos # 1 CPU 1 dev w. route DoS
141 pktgen.conf-1-1-ip6 # 1 CPU 1 dev ipv6
142 pktgen.conf-1-1-ip6-rdos # 1 CPU 1 dev ipv6 w. route DoS
143 pktgen.conf-1-1-flows # 1 CPU 1 dev multiple flows.
145 Run in shell: ./pktgen.conf-X-Y It does all the setup including sending.
148 Interrupt affinity
149 ===================
150 Note when adding devices to a specific CPU there good idea to also assign
151 /proc/irq/XX/smp_affinity so the TX-interrupts gets bound to the same CPU.
152 as this reduces cache bouncing when freeing skb's.
155 Current commands and configuration options
156 ==========================================
158 ** Pgcontrol commands:
160 start
161 stop
163 ** Thread commands:
165 add_device
166 rem_device_all
167 max_before_softirq
170 ** Device commands:
172 count
173 clone_skb
174 debug
176 frags
177 delay
179 src_mac_count
180 dst_mac_count
182 pkt_size
183 min_pkt_size
184 max_pkt_size
186 mpls
188 udp_src_min
189 udp_src_max
191 udp_dst_min
192 udp_dst_max
194 flag
195 IPSRC_RND
196 TXSIZE_RND
197 IPDST_RND
198 UDPSRC_RND
199 UDPDST_RND
200 MACSRC_RND
201 MACDST_RND
203 dst_min
204 dst_max
206 src_min
207 src_max
209 dst_mac
210 src_mac
212 clear_counters
214 dst6
215 src6
217 flows
218 flowlen
220 References:
221 ftp://robur.slu.se/pub/Linux/net-development/pktgen-testing/
222 ftp://robur.slu.se/pub/Linux/net-development/pktgen-testing/examples/
224 Paper from Linux-Kongress in Erlangen 2004.
225 ftp://robur.slu.se/pub/Linux/net-development/pktgen-testing/pktgen_paper.pdf
227 Thanks to:
228 Grant Grundler for testing on IA-64 and parisc, Harald Welte, Lennert Buytenhek
229 Stephen Hemminger, Andi Kleen, Dave Miller and many others.
232 Good luck with the linux net-development.