7 cgroup subsys "blkio" implements the block io controller. There seems to be
8 a need of various kinds of IO control policies (like proportional BW, max BW)
9 both at leaf nodes as well as at intermediate nodes in a storage hierarchy.
10 Plan is to use the same cgroup based management interface for blkio controller
11 and based on user options switch IO policies in the background.
13 One IO control policy is throttling policy which can be used to
14 specify upper IO rate limits on devices. This policy is implemented in
15 generic block layer and can be used on leaf nodes as well as higher
16 level logical devices like device mapper.
21 Throttling/Upper Limit policy
22 -----------------------------
23 Enable Block IO controller::
27 Enable throttling in block layer::
29 CONFIG_BLK_DEV_THROTTLING=y
31 Mount blkio controller (see cgroups.txt, Why are cgroups needed?)::
33 mount -t cgroup -o blkio none /sys/fs/cgroup/blkio
35 Specify a bandwidth rate on particular device for root group. The format
36 for policy is "<major>:<minor> <bytes_per_second>"::
38 echo "8:16 1048576" > /sys/fs/cgroup/blkio/blkio.throttle.read_bps_device
40 This will put a limit of 1MB/second on reads happening for root group
41 on device having major/minor number 8:16.
43 Run dd to read a file and see if rate is throttled to 1MB/s or not::
45 # dd iflag=direct if=/mnt/common/zerofile of=/dev/null bs=4K count=1024
48 4194304 bytes (4.2 MB) copied, 4.0001 s, 1.0 MB/s
50 Limits for writes can be put using blkio.throttle.write_bps_device file.
55 Throttling implements hierarchy support; however,
56 throttling's hierarchy support is enabled iff "sane_behavior" is
57 enabled from cgroup side, which currently is a development option and
58 not publicly available.
60 If somebody created a hierarchy like as follows::
68 Throttling with "sane_behavior" will handle the
69 hierarchy correctly. For throttling, all limits apply
70 to the whole subtree while all statistics are local to the IOs
71 directly generated by tasks in that cgroup.
73 Throttling without "sane_behavior" enabled from cgroup side will
74 practically treat all groups at same level as if it looks like the
79 root test1 test2 test3
81 Various user visible config options
82 ===================================
87 CONFIG_BFQ_CGROUP_DEBUG
88 Debug help. Right now some additional stats file show up in cgroup
89 if this option is enabled.
91 CONFIG_BLK_DEV_THROTTLING
92 Enable block device throttling support in block layer.
94 Details of cgroup files
95 =======================
97 Proportional weight policy files
98 --------------------------------
101 Specifies per cgroup weight. This is default weight of the group
102 on all the devices until and unless overridden by per device rule.
103 (See blkio.weight_device).
104 Currently allowed range of weights is from 10 to 1000.
107 One can specify per cgroup per device rules using this interface.
108 These rules override the default value of group weight as specified
111 Following is the format::
113 # echo dev_maj:dev_minor weight > blkio.weight_device
115 Configure weight=300 on /dev/sdb (8:16) in this cgroup::
117 # echo 8:16 300 > blkio.weight_device
118 # cat blkio.weight_device
122 Configure weight=500 on /dev/sda (8:0) in this cgroup::
124 # echo 8:0 500 > blkio.weight_device
125 # cat blkio.weight_device
130 Remove specific weight for /dev/sda in this cgroup::
132 # echo 8:0 0 > blkio.weight_device
133 # cat blkio.weight_device
138 Disk time allocated to cgroup per device in milliseconds. First
139 two fields specify the major and minor number of the device and
140 third field specifies the disk time allocated to group in
144 Number of sectors transferred to/from disk by the group. First
145 two fields specify the major and minor number of the device and
146 third field specifies the number of sectors transferred by the
147 group to/from the device.
149 blkio.io_service_bytes
150 Number of bytes transferred to/from the disk by the group. These
151 are further divided by the type of operation - read or write, sync
152 or async. First two fields specify the major and minor number of the
153 device, third field specifies the operation type and the fourth field
154 specifies the number of bytes.
157 Number of IOs (bio) issued to the disk by the group. These
158 are further divided by the type of operation - read or write, sync
159 or async. First two fields specify the major and minor number of the
160 device, third field specifies the operation type and the fourth field
161 specifies the number of IOs.
163 blkio.io_service_time
164 Total amount of time between request dispatch and request completion
165 for the IOs done by this cgroup. This is in nanoseconds to make it
166 meaningful for flash devices too. For devices with queue depth of 1,
167 this time represents the actual service time. When queue_depth > 1,
168 that is no longer true as requests may be served out of order. This
169 may cause the service time for a given IO to include the service time
170 of multiple IOs when served out of order which may result in total
171 io_service_time > actual time elapsed. This time is further divided by
172 the type of operation - read or write, sync or async. First two fields
173 specify the major and minor number of the device, third field
174 specifies the operation type and the fourth field specifies the
175 io_service_time in ns.
178 Total amount of time the IOs for this cgroup spent waiting in the
179 scheduler queues for service. This can be greater than the total time
180 elapsed since it is cumulative io_wait_time for all IOs. It is not a
181 measure of total time the cgroup spent waiting but rather a measure of
182 the wait_time for its individual IOs. For devices with queue_depth > 1
183 this metric does not include the time spent waiting for service once
184 the IO is dispatched to the device but till it actually gets serviced
185 (there might be a time lag here due to re-ordering of requests by the
186 device). This is in nanoseconds to make it meaningful for flash
187 devices too. This time is further divided by the type of operation -
188 read or write, sync or async. First two fields specify the major and
189 minor number of the device, third field specifies the operation type
190 and the fourth field specifies the io_wait_time in ns.
193 Total number of bios/requests merged into requests belonging to this
194 cgroup. This is further divided by the type of operation - read or
195 write, sync or async.
198 Total number of requests queued up at any given instant for this
199 cgroup. This is further divided by the type of operation - read or
200 write, sync or async.
203 Debugging aid only enabled if CONFIG_BFQ_CGROUP_DEBUG=y.
204 The average queue size for this cgroup over the entire time of this
205 cgroup's existence. Queue size samples are taken each time one of the
206 queues of this cgroup gets a timeslice.
208 blkio.group_wait_time
209 Debugging aid only enabled if CONFIG_BFQ_CGROUP_DEBUG=y.
210 This is the amount of time the cgroup had to wait since it became busy
211 (i.e., went from 0 to 1 request queued) to get a timeslice for one of
212 its queues. This is different from the io_wait_time which is the
213 cumulative total of the amount of time spent by each IO in that cgroup
214 waiting in the scheduler queue. This is in nanoseconds. If this is
215 read when the cgroup is in a waiting (for timeslice) state, the stat
216 will only report the group_wait_time accumulated till the last time it
217 got a timeslice and will not include the current delta.
220 Debugging aid only enabled if CONFIG_BFQ_CGROUP_DEBUG=y.
221 This is the amount of time a cgroup spends without any pending
222 requests when not being served, i.e., it does not include any time
223 spent idling for one of the queues of the cgroup. This is in
224 nanoseconds. If this is read when the cgroup is in an empty state,
225 the stat will only report the empty_time accumulated till the last
226 time it had a pending request and will not include the current delta.
229 Debugging aid only enabled if CONFIG_BFQ_CGROUP_DEBUG=y.
230 This is the amount of time spent by the IO scheduler idling for a
231 given cgroup in anticipation of a better request than the existing ones
232 from other queues/cgroups. This is in nanoseconds. If this is read
233 when the cgroup is in an idling state, the stat will only report the
234 idle_time accumulated till the last idle period and will not include
238 Debugging aid only enabled if CONFIG_BFQ_CGROUP_DEBUG=y. This
239 gives the statistics about how many a times a group was dequeued
240 from service tree of the device. First two fields specify the major
241 and minor number of the device and third field specifies the number
242 of times a group was dequeued from a particular device.
245 Recursive version of various stats. These files show the
246 same information as their non-recursive counterparts but
247 include stats from all the descendant cgroups.
249 Throttling/Upper limit policy files
250 -----------------------------------
251 blkio.throttle.read_bps_device
252 Specifies upper limit on READ rate from the device. IO rate is
253 specified in bytes per second. Rules are per device. Following is
256 echo "<major>:<minor> <rate_bytes_per_second>" > /cgrp/blkio.throttle.read_bps_device
258 blkio.throttle.write_bps_device
259 Specifies upper limit on WRITE rate to the device. IO rate is
260 specified in bytes per second. Rules are per device. Following is
263 echo "<major>:<minor> <rate_bytes_per_second>" > /cgrp/blkio.throttle.write_bps_device
265 blkio.throttle.read_iops_device
266 Specifies upper limit on READ rate from the device. IO rate is
267 specified in IO per second. Rules are per device. Following is
270 echo "<major>:<minor> <rate_io_per_second>" > /cgrp/blkio.throttle.read_iops_device
272 blkio.throttle.write_iops_device
273 Specifies upper limit on WRITE rate to the device. IO rate is
274 specified in io per second. Rules are per device. Following is
277 echo "<major>:<minor> <rate_io_per_second>" > /cgrp/blkio.throttle.write_iops_device
279 Note: If both BW and IOPS rules are specified for a device, then IO is
280 subjected to both the constraints.
282 blkio.throttle.io_serviced
283 Number of IOs (bio) issued to the disk by the group. These
284 are further divided by the type of operation - read or write, sync
285 or async. First two fields specify the major and minor number of the
286 device, third field specifies the operation type and the fourth field
287 specifies the number of IOs.
289 blkio.throttle.io_service_bytes
290 Number of bytes transferred to/from the disk by the group. These
291 are further divided by the type of operation - read or write, sync
292 or async. First two fields specify the major and minor number of the
293 device, third field specifies the operation type and the fourth field
294 specifies the number of bytes.
296 Common files among various policies
297 -----------------------------------
299 Writing an int to this file will result in resetting all the stats