view drivers/md/Kconfig @ 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
line source
1 #
2 # Block device driver configuration
3 #
5 menu "Multi-device support (RAID and LVM)"
7 config MD
8 bool "Multiple devices driver support (RAID and LVM)"
9 help
10 Support multiple physical spindles through a single logical device.
11 Required for RAID and logical volume management.
13 config BLK_DEV_MD
14 tristate "RAID support"
15 depends on MD
16 ---help---
17 This driver lets you combine several hard disk partitions into one
18 logical block device. This can be used to simply append one
19 partition to another one or to combine several redundant hard disks
20 into a RAID1/4/5 device so as to provide protection against hard
21 disk failures. This is called "Software RAID" since the combining of
22 the partitions is done by the kernel. "Hardware RAID" means that the
23 combining is done by a dedicated controller; if you have such a
24 controller, you do not need to say Y here.
26 More information about Software RAID on Linux is contained in the
27 Software RAID mini-HOWTO, available from
28 <http://www.tldp.org/docs.html#howto>. There you will also learn
29 where to get the supporting user space utilities raidtools.
31 If unsure, say N.
33 config MD_LINEAR
34 tristate "Linear (append) mode"
35 depends on BLK_DEV_MD
36 ---help---
37 If you say Y here, then your multiple devices driver will be able to
38 use the so-called linear mode, i.e. it will combine the hard disk
39 partitions by simply appending one to the other.
41 To compile this as a module, choose M here: the module
42 will be called linear.
44 If unsure, say Y.
46 config MD_RAID0
47 tristate "RAID-0 (striping) mode"
48 depends on BLK_DEV_MD
49 ---help---
50 If you say Y here, then your multiple devices driver will be able to
51 use the so-called raid0 mode, i.e. it will combine the hard disk
52 partitions into one logical device in such a fashion as to fill them
53 up evenly, one chunk here and one chunk there. This will increase
54 the throughput rate if the partitions reside on distinct disks.
56 Information about Software RAID on Linux is contained in the
57 Software-RAID mini-HOWTO, available from
58 <http://www.tldp.org/docs.html#howto>. There you will also
59 learn where to get the supporting user space utilities raidtools.
61 To compile this as a module, choose M here: the module
62 will be called raid0.
64 If unsure, say Y.
66 config MD_RAID1
67 tristate "RAID-1 (mirroring) mode"
68 depends on BLK_DEV_MD
69 ---help---
70 A RAID-1 set consists of several disk drives which are exact copies
71 of each other. In the event of a mirror failure, the RAID driver
72 will continue to use the operational mirrors in the set, providing
73 an error free MD (multiple device) to the higher levels of the
74 kernel. In a set with N drives, the available space is the capacity
75 of a single drive, and the set protects against a failure of (N - 1)
76 drives.
78 Information about Software RAID on Linux is contained in the
79 Software-RAID mini-HOWTO, available from
80 <http://www.tldp.org/docs.html#howto>. There you will also
81 learn where to get the supporting user space utilities raidtools.
83 If you want to use such a RAID-1 set, say Y. To compile this code
84 as a module, choose M here: the module will be called raid1.
86 If unsure, say Y.
88 config MD_RAID10
89 tristate "RAID-10 (mirrored striping) mode (EXPERIMENTAL)"
90 depends on BLK_DEV_MD && EXPERIMENTAL
91 ---help---
92 RAID-10 provides a combination of striping (RAID-0) and
93 mirroring (RAID-1) with easier configuration and more flexible
94 layout.
95 Unlike RAID-0, but like RAID-1, RAID-10 requires all devices to
96 be the same size (or at least, only as much as the smallest device
97 will be used).
98 RAID-10 provides a variety of layouts that provide different levels
99 of redundancy and performance.
101 RAID-10 requires mdadm-1.7.0 or later, available at:
103 ftp://ftp.kernel.org/pub/linux/utils/raid/mdadm/
105 If unsure, say Y.
107 config MD_RAID456
108 tristate "RAID-4/RAID-5/RAID-6 mode"
109 depends on BLK_DEV_MD
110 ---help---
111 A RAID-5 set of N drives with a capacity of C MB per drive provides
112 the capacity of C * (N - 1) MB, and protects against a failure
113 of a single drive. For a given sector (row) number, (N - 1) drives
114 contain data sectors, and one drive contains the parity protection.
115 For a RAID-4 set, the parity blocks are present on a single drive,
116 while a RAID-5 set distributes the parity across the drives in one
117 of the available parity distribution methods.
119 A RAID-6 set of N drives with a capacity of C MB per drive
120 provides the capacity of C * (N - 2) MB, and protects
121 against a failure of any two drives. For a given sector
122 (row) number, (N - 2) drives contain data sectors, and two
123 drives contains two independent redundancy syndromes. Like
124 RAID-5, RAID-6 distributes the syndromes across the drives
125 in one of the available parity distribution methods.
127 Information about Software RAID on Linux is contained in the
128 Software-RAID mini-HOWTO, available from
129 <http://www.tldp.org/docs.html#howto>. There you will also
130 learn where to get the supporting user space utilities raidtools.
132 If you want to use such a RAID-4/RAID-5/RAID-6 set, say Y. To
133 compile this code as a module, choose M here: the module
134 will be called raid456.
136 If unsure, say Y.
138 config MD_RAID5_RESHAPE
139 bool "Support adding drives to a raid-5 array (experimental)"
140 depends on MD_RAID456 && EXPERIMENTAL
141 ---help---
142 A RAID-5 set can be expanded by adding extra drives. This
143 requires "restriping" the array which means (almost) every
144 block must be written to a different place.
146 This option allows such restriping to be done while the array
147 is online. However it is still EXPERIMENTAL code. It should
148 work, but please be sure that you have backups.
150 You will need mdadm version 2.4.1 or later to use this
151 feature safely. During the early stage of reshape there is
152 a critical section where live data is being over-written. A
153 crash during this time needs extra care for recovery. The
154 newer mdadm takes a copy of the data in the critical section
155 and will restore it, if necessary, after a crash.
157 The mdadm usage is e.g.
158 mdadm --grow /dev/md1 --raid-disks=6
159 to grow '/dev/md1' to having 6 disks.
161 Note: The array can only be expanded, not contracted.
162 There should be enough spares already present to make the new
163 array workable.
165 config MD_MULTIPATH
166 tristate "Multipath I/O support"
167 depends on BLK_DEV_MD
168 help
169 Multipath-IO is the ability of certain devices to address the same
170 physical disk over multiple 'IO paths'. The code ensures that such
171 paths can be defined and handled at runtime, and ensures that a
172 transparent failover to the backup path(s) happens if a IO errors
173 arrives on the primary path.
175 If unsure, say N.
177 config MD_FAULTY
178 tristate "Faulty test module for MD"
179 depends on BLK_DEV_MD
180 help
181 The "faulty" module allows for a block device that occasionally returns
182 read or write errors. It is useful for testing.
184 In unsure, say N.
186 config BLK_DEV_DM
187 tristate "Device mapper support"
188 depends on MD
189 ---help---
190 Device-mapper is a low level volume manager. It works by allowing
191 people to specify mappings for ranges of logical sectors. Various
192 mapping types are available, in addition people may write their own
193 modules containing custom mappings if they wish.
195 Higher level volume managers such as LVM2 use this driver.
197 To compile this as a module, choose M here: the module will be
198 called dm-mod.
200 If unsure, say N.
202 config DM_CRYPT
203 tristate "Crypt target support"
204 depends on BLK_DEV_DM && EXPERIMENTAL
205 select CRYPTO
206 ---help---
207 This device-mapper target allows you to create a device that
208 transparently encrypts the data on it. You'll need to activate
209 the ciphers you're going to use in the cryptoapi configuration.
211 Information on how to use dm-crypt can be found on
213 <http://www.saout.de/misc/dm-crypt/>
215 To compile this code as a module, choose M here: the module will
216 be called dm-crypt.
218 If unsure, say N.
220 config DM_SNAPSHOT
221 tristate "Snapshot target (EXPERIMENTAL)"
222 depends on BLK_DEV_DM && EXPERIMENTAL
223 ---help---
224 Allow volume managers to take writable snapshots of a device.
226 config DM_MIRROR
227 tristate "Mirror target (EXPERIMENTAL)"
228 depends on BLK_DEV_DM && EXPERIMENTAL
229 ---help---
230 Allow volume managers to mirror logical volumes, also
231 needed for live data migration tools such as 'pvmove'.
233 config DM_ZERO
234 tristate "Zero target (EXPERIMENTAL)"
235 depends on BLK_DEV_DM && EXPERIMENTAL
236 ---help---
237 A target that discards writes, and returns all zeroes for
238 reads. Useful in some recovery situations.
240 config DM_MULTIPATH
241 tristate "Multipath target (EXPERIMENTAL)"
242 depends on BLK_DEV_DM && EXPERIMENTAL
243 ---help---
244 Allow volume managers to support multipath hardware.
247 tristate "EMC CX/AX multipath support (EXPERIMENTAL)"
249 ---help---
250 Multipath support for EMC CX/AX series hardware.
252 endmenu