1 ============================ 1 ============================
2 A block layer cache (bcache) 2 A block layer cache (bcache)
3 ============================ 3 ============================
4 4
5 Say you've got a big slow raid 6, and an ssd o 5 Say you've got a big slow raid 6, and an ssd or three. Wouldn't it be
6 nice if you could use them as cache... Hence b 6 nice if you could use them as cache... Hence bcache.
7 7
8 The bcache wiki can be found at: !! 8 Wiki and git repositories are at:
9 https://bcache.evilpiepirate.org <<
10 9
11 This is the git repository of bcache-tools: !! 10 - https://bcache.evilpiepirate.org
12 https://git.kernel.org/pub/scm/linux/kernel/ !! 11 - http://evilpiepirate.org/git/linux-bcache.git
13 !! 12 - https://evilpiepirate.org/git/bcache-tools.git
14 The latest bcache kernel code can be found fro <<
15 https://git.kernel.org/pub/scm/linux/kernel/ <<
16 13
17 It's designed around the performance character 14 It's designed around the performance characteristics of SSDs - it only allocates
18 in erase block sized buckets, and it uses a hy 15 in erase block sized buckets, and it uses a hybrid btree/log to track cached
19 extents (which can be anywhere from a single s 16 extents (which can be anywhere from a single sector to the bucket size). It's
20 designed to avoid random writes at all costs; 17 designed to avoid random writes at all costs; it fills up an erase block
21 sequentially, then issues a discard before reu 18 sequentially, then issues a discard before reusing it.
22 19
23 Both writethrough and writeback caching are su 20 Both writethrough and writeback caching are supported. Writeback defaults to
24 off, but can be switched on and off arbitraril 21 off, but can be switched on and off arbitrarily at runtime. Bcache goes to
25 great lengths to protect your data - it reliab 22 great lengths to protect your data - it reliably handles unclean shutdown. (It
26 doesn't even have a notion of a clean shutdown 23 doesn't even have a notion of a clean shutdown; bcache simply doesn't return
27 writes as completed until they're on stable st 24 writes as completed until they're on stable storage).
28 25
29 Writeback caching can use most of the cache fo 26 Writeback caching can use most of the cache for buffering writes - writing
30 dirty data to the backing device is always don 27 dirty data to the backing device is always done sequentially, scanning from the
31 start to the end of the index. 28 start to the end of the index.
32 29
33 Since random IO is what SSDs excel at, there g 30 Since random IO is what SSDs excel at, there generally won't be much benefit
34 to caching large sequential IO. Bcache detects 31 to caching large sequential IO. Bcache detects sequential IO and skips it;
35 it also keeps a rolling average of the IO size 32 it also keeps a rolling average of the IO sizes per task, and as long as the
36 average is above the cutoff it will skip all I 33 average is above the cutoff it will skip all IO from that task - instead of
37 caching the first 512k after every seek. Backu 34 caching the first 512k after every seek. Backups and large file copies should
38 thus entirely bypass the cache. 35 thus entirely bypass the cache.
39 36
40 In the event of a data IO error on the flash i 37 In the event of a data IO error on the flash it will try to recover by reading
41 from disk or invalidating cache entries. For 38 from disk or invalidating cache entries. For unrecoverable errors (meta data
42 or dirty data), caching is automatically disab 39 or dirty data), caching is automatically disabled; if dirty data was present
43 in the cache it first disables writeback cachi 40 in the cache it first disables writeback caching and waits for all dirty data
44 to be flushed. 41 to be flushed.
45 42
46 Getting started: 43 Getting started:
47 You'll need bcache util from the bcache-tools !! 44 You'll need make-bcache from the bcache-tools repository. Both the cache device
48 and backing device must be formatted before us 45 and backing device must be formatted before use::
49 46
50 bcache make -B /dev/sdb !! 47 make-bcache -B /dev/sdb
51 bcache make -C /dev/sdc !! 48 make-bcache -C /dev/sdc
52 49
53 `bcache make` has the ability to format multip !! 50 make-bcache has the ability to format multiple devices at the same time - if
54 you format your backing devices and cache devi 51 you format your backing devices and cache device at the same time, you won't
55 have to manually attach:: 52 have to manually attach::
56 53
57 bcache make -B /dev/sda /dev/sdb -C /dev/sdc !! 54 make-bcache -B /dev/sda /dev/sdb -C /dev/sdc
58 <<
59 If your bcache-tools is not updated to latest <<
60 unified `bcache` utility, you may use the lega <<
61 bcache device with same -B and -C parameters. <<
62 55
63 bcache-tools now ships udev rules, and bcache 56 bcache-tools now ships udev rules, and bcache devices are known to the kernel
64 immediately. Without udev, you can manually r 57 immediately. Without udev, you can manually register devices like this::
65 58
66 echo /dev/sdb > /sys/fs/bcache/register 59 echo /dev/sdb > /sys/fs/bcache/register
67 echo /dev/sdc > /sys/fs/bcache/register 60 echo /dev/sdc > /sys/fs/bcache/register
68 61
69 Registering the backing device makes the bcach 62 Registering the backing device makes the bcache device show up in /dev; you can
70 now format it and use it as normal. But the fi 63 now format it and use it as normal. But the first time using a new bcache
71 device, it'll be running in passthrough mode u 64 device, it'll be running in passthrough mode until you attach it to a cache.
72 If you are thinking about using bcache later, 65 If you are thinking about using bcache later, it is recommended to setup all your
73 slow devices as bcache backing devices without 66 slow devices as bcache backing devices without a cache, and you can choose to add
74 a caching device later. 67 a caching device later.
75 See 'ATTACHING' section below. 68 See 'ATTACHING' section below.
76 69
77 The devices show up as:: 70 The devices show up as::
78 71
79 /dev/bcache<N> 72 /dev/bcache<N>
80 73
81 As well as (with udev):: 74 As well as (with udev)::
82 75
83 /dev/bcache/by-uuid/<uuid> 76 /dev/bcache/by-uuid/<uuid>
84 /dev/bcache/by-label/<label> 77 /dev/bcache/by-label/<label>
85 78
86 To get started:: 79 To get started::
87 80
88 mkfs.ext4 /dev/bcache0 81 mkfs.ext4 /dev/bcache0
89 mount /dev/bcache0 /mnt 82 mount /dev/bcache0 /mnt
90 83
91 You can control bcache devices through sysfs a 84 You can control bcache devices through sysfs at /sys/block/bcache<N>/bcache .
92 You can also control them through /sys/fs//bca 85 You can also control them through /sys/fs//bcache/<cset-uuid>/ .
93 86
94 Cache devices are managed as sets; multiple ca 87 Cache devices are managed as sets; multiple caches per set isn't supported yet
95 but will allow for mirroring of metadata and d 88 but will allow for mirroring of metadata and dirty data in the future. Your new
96 cache set shows up as /sys/fs/bcache/<UUID> 89 cache set shows up as /sys/fs/bcache/<UUID>
97 90
98 Attaching 91 Attaching
99 --------- 92 ---------
100 93
101 After your cache device and backing device are 94 After your cache device and backing device are registered, the backing device
102 must be attached to your cache set to enable c 95 must be attached to your cache set to enable caching. Attaching a backing
103 device to a cache set is done thusly, with the 96 device to a cache set is done thusly, with the UUID of the cache set in
104 /sys/fs/bcache:: 97 /sys/fs/bcache::
105 98
106 echo <CSET-UUID> > /sys/block/bcache0/bcache 99 echo <CSET-UUID> > /sys/block/bcache0/bcache/attach
107 100
108 This only has to be done once. The next time y 101 This only has to be done once. The next time you reboot, just reregister all
109 your bcache devices. If a backing device has d 102 your bcache devices. If a backing device has data in a cache somewhere, the
110 /dev/bcache<N> device won't be created until t 103 /dev/bcache<N> device won't be created until the cache shows up - particularly
111 important if you have writeback caching turned 104 important if you have writeback caching turned on.
112 105
113 If you're booting up and your cache device is 106 If you're booting up and your cache device is gone and never coming back, you
114 can force run the backing device:: 107 can force run the backing device::
115 108
116 echo 1 > /sys/block/sdb/bcache/running 109 echo 1 > /sys/block/sdb/bcache/running
117 110
118 (You need to use /sys/block/sdb (or whatever y 111 (You need to use /sys/block/sdb (or whatever your backing device is called), not
119 /sys/block/bcache0, because bcache0 doesn't ex 112 /sys/block/bcache0, because bcache0 doesn't exist yet. If you're using a
120 partition, the bcache directory would be at /s 113 partition, the bcache directory would be at /sys/block/sdb/sdb2/bcache)
121 114
122 The backing device will still use that cache s 115 The backing device will still use that cache set if it shows up in the future,
123 but all the cached data will be invalidated. I 116 but all the cached data will be invalidated. If there was dirty data in the
124 cache, don't expect the filesystem to be recov 117 cache, don't expect the filesystem to be recoverable - you will have massive
125 filesystem corruption, though ext4's fsck does 118 filesystem corruption, though ext4's fsck does work miracles.
126 119
127 Error Handling 120 Error Handling
128 -------------- 121 --------------
129 122
130 Bcache tries to transparently handle IO errors 123 Bcache tries to transparently handle IO errors to/from the cache device without
131 affecting normal operation; if it sees too man 124 affecting normal operation; if it sees too many errors (the threshold is
132 configurable, and defaults to 0) it shuts down 125 configurable, and defaults to 0) it shuts down the cache device and switches all
133 the backing devices to passthrough mode. 126 the backing devices to passthrough mode.
134 127
135 - For reads from the cache, if they error we 128 - For reads from the cache, if they error we just retry the read from the
136 backing device. 129 backing device.
137 130
138 - For writethrough writes, if the write to th 131 - For writethrough writes, if the write to the cache errors we just switch to
139 invalidating the data at that lba in the ca 132 invalidating the data at that lba in the cache (i.e. the same thing we do for
140 a write that bypasses the cache) 133 a write that bypasses the cache)
141 134
142 - For writeback writes, we currently pass tha 135 - For writeback writes, we currently pass that error back up to the
143 filesystem/userspace. This could be improve 136 filesystem/userspace. This could be improved - we could retry it as a write
144 that skips the cache so we don't have to er 137 that skips the cache so we don't have to error the write.
145 138
146 - When we detach, we first try to flush any d 139 - When we detach, we first try to flush any dirty data (if we were running in
147 writeback mode). It currently doesn't do an 140 writeback mode). It currently doesn't do anything intelligent if it fails to
148 read some of the dirty data, though. 141 read some of the dirty data, though.
149 142
150 143
151 Howto/cookbook 144 Howto/cookbook
152 -------------- 145 --------------
153 146
154 A) Starting a bcache with a missing caching de 147 A) Starting a bcache with a missing caching device
155 148
156 If registering the backing device doesn't help 149 If registering the backing device doesn't help, it's already there, you just need
157 to force it to run without the cache:: 150 to force it to run without the cache::
158 151
159 host:~# echo /dev/sdb1 > /sys/fs/bcach 152 host:~# echo /dev/sdb1 > /sys/fs/bcache/register
160 [ 119.844831] bcache: register_bcache 153 [ 119.844831] bcache: register_bcache() error opening /dev/sdb1: device already registered
161 154
162 Next, you try to register your caching device 155 Next, you try to register your caching device if it's present. However
163 if it's absent, or registration fails for some 156 if it's absent, or registration fails for some reason, you can still
164 start your bcache without its cache, like so:: 157 start your bcache without its cache, like so::
165 158
166 host:/sys/block/sdb/sdb1/bcache# echo 159 host:/sys/block/sdb/sdb1/bcache# echo 1 > running
167 160
168 Note that this may cause data loss if you were 161 Note that this may cause data loss if you were running in writeback mode.
169 162
170 163
171 B) Bcache does not find its cache:: 164 B) Bcache does not find its cache::
172 165
173 host:/sys/block/md5/bcache# echo 02265 166 host:/sys/block/md5/bcache# echo 0226553a-37cf-41d5-b3ce-8b1e944543a8 > attach
174 [ 1933.455082] bcache: bch_cached_dev_ 167 [ 1933.455082] bcache: bch_cached_dev_attach() Couldn't find uuid for md5 in set
175 [ 1933.478179] bcache: __cached_dev_st 168 [ 1933.478179] bcache: __cached_dev_store() Can't attach 0226553a-37cf-41d5-b3ce-8b1e944543a8
176 [ 1933.478179] : cache set not found 169 [ 1933.478179] : cache set not found
177 170
178 In this case, the caching device was simply no 171 In this case, the caching device was simply not registered at boot
179 or disappeared and came back, and needs to be 172 or disappeared and came back, and needs to be (re-)registered::
180 173
181 host:/sys/block/md5/bcache# echo /dev/ 174 host:/sys/block/md5/bcache# echo /dev/sdh2 > /sys/fs/bcache/register
182 175
183 176
184 C) Corrupt bcache crashes the kernel at device 177 C) Corrupt bcache crashes the kernel at device registration time:
185 178
186 This should never happen. If it does happen, 179 This should never happen. If it does happen, then you have found a bug!
187 Please report it to the bcache development lis 180 Please report it to the bcache development list: linux-bcache@vger.kernel.org
188 181
189 Be sure to provide as much information that yo 182 Be sure to provide as much information that you can including kernel dmesg
190 output if available so that we may assist. 183 output if available so that we may assist.
191 184
192 185
193 D) Recovering data without bcache: 186 D) Recovering data without bcache:
194 187
195 If bcache is not available in the kernel, a fi 188 If bcache is not available in the kernel, a filesystem on the backing
196 device is still available at an 8KiB offset. S 189 device is still available at an 8KiB offset. So either via a loopdev
197 of the backing device created with --offset 8K 190 of the backing device created with --offset 8K, or any value defined by
198 --data-offset when you originally formatted bc !! 191 --data-offset when you originally formatted bcache with `make-bcache`.
199 192
200 For example:: 193 For example::
201 194
202 losetup -o 8192 /dev/loop0 /dev/your_b 195 losetup -o 8192 /dev/loop0 /dev/your_bcache_backing_dev
203 196
204 This should present your unmodified backing de 197 This should present your unmodified backing device data in /dev/loop0
205 198
206 If your cache is in writethrough mode, then yo 199 If your cache is in writethrough mode, then you can safely discard the
207 cache device without losing data. !! 200 cache device without loosing data.
208 201
209 202
210 E) Wiping a cache device 203 E) Wiping a cache device
211 204
212 :: 205 ::
213 206
214 host:~# wipefs -a /dev/sdh2 207 host:~# wipefs -a /dev/sdh2
215 16 bytes were erased at offset 0x1018 208 16 bytes were erased at offset 0x1018 (bcache)
216 they were: c6 85 73 f6 4e 1a 45 ca 82 209 they were: c6 85 73 f6 4e 1a 45 ca 82 65 f5 7f 48 ba 6d 81
217 210
218 After you boot back with bcache enabled, you r 211 After you boot back with bcache enabled, you recreate the cache and attach it::
219 212
220 host:~# bcache make -C /dev/sdh2 !! 213 host:~# make-bcache -C /dev/sdh2
221 UUID: 7be7e175-8f4c- 214 UUID: 7be7e175-8f4c-4f99-94b2-9c904d227045
222 Set UUID: 5bc072a8-ab17- 215 Set UUID: 5bc072a8-ab17-446d-9744-e247949913c1
223 version: 0 216 version: 0
224 nbuckets: 106874 217 nbuckets: 106874
225 block_size: 1 218 block_size: 1
226 bucket_size: 1024 219 bucket_size: 1024
227 nr_in_set: 1 220 nr_in_set: 1
228 nr_this_dev: 0 221 nr_this_dev: 0
229 first_bucket: 1 222 first_bucket: 1
230 [ 650.511912] bcache: run_cache_set() 223 [ 650.511912] bcache: run_cache_set() invalidating existing data
231 [ 650.549228] bcache: register_cache( 224 [ 650.549228] bcache: register_cache() registered cache device sdh2
232 225
233 start backing device with missing cache:: 226 start backing device with missing cache::
234 227
235 host:/sys/block/md5/bcache# echo 1 > r 228 host:/sys/block/md5/bcache# echo 1 > running
236 229
237 attach new cache:: 230 attach new cache::
238 231
239 host:/sys/block/md5/bcache# echo 5bc07 232 host:/sys/block/md5/bcache# echo 5bc072a8-ab17-446d-9744-e247949913c1 > attach
240 [ 865.276616] bcache: bch_cached_dev_ 233 [ 865.276616] bcache: bch_cached_dev_attach() Caching md5 as bcache0 on set 5bc072a8-ab17-446d-9744-e247949913c1
241 234
242 235
243 F) Remove or replace a caching device:: 236 F) Remove or replace a caching device::
244 237
245 host:/sys/block/sda/sda7/bcache# echo 238 host:/sys/block/sda/sda7/bcache# echo 1 > detach
246 [ 695.872542] bcache: cached_dev_deta 239 [ 695.872542] bcache: cached_dev_detach_finish() Caching disabled for sda7
247 240
248 host:~# wipefs -a /dev/nvme0n1p4 241 host:~# wipefs -a /dev/nvme0n1p4
249 wipefs: error: /dev/nvme0n1p4: probing 242 wipefs: error: /dev/nvme0n1p4: probing initialization failed: Device or resource busy
250 Ooops, it's disabled, but not unregist 243 Ooops, it's disabled, but not unregistered, so it's still protected
251 244
252 We need to go and unregister it:: 245 We need to go and unregister it::
253 246
254 host:/sys/fs/bcache/b7ba27a1-2398-4649 247 host:/sys/fs/bcache/b7ba27a1-2398-4649-8ae3-0959f57ba128# ls -l cache0
255 lrwxrwxrwx 1 root root 0 Feb 25 18:33 248 lrwxrwxrwx 1 root root 0 Feb 25 18:33 cache0 -> ../../../devices/pci0000:00/0000:00:1d.0/0000:70:00.0/nvme/nvme0/nvme0n1/nvme0n1p4/bcache/
256 host:/sys/fs/bcache/b7ba27a1-2398-4649 249 host:/sys/fs/bcache/b7ba27a1-2398-4649-8ae3-0959f57ba128# echo 1 > stop
257 kernel: [ 917.041908] bcache: cache_s 250 kernel: [ 917.041908] bcache: cache_set_free() Cache set b7ba27a1-2398-4649-8ae3-0959f57ba128 unregistered
258 251
259 Now we can wipe it:: 252 Now we can wipe it::
260 253
261 host:~# wipefs -a /dev/nvme0n1p4 254 host:~# wipefs -a /dev/nvme0n1p4
262 /dev/nvme0n1p4: 16 bytes were erased a 255 /dev/nvme0n1p4: 16 bytes were erased at offset 0x00001018 (bcache): c6 85 73 f6 4e 1a 45 ca 82 65 f5 7f 48 ba 6d 81
263 256
264 257
265 G) dm-crypt and bcache 258 G) dm-crypt and bcache
266 259
267 First setup bcache unencrypted and then instal 260 First setup bcache unencrypted and then install dmcrypt on top of
268 /dev/bcache<N> This will work faster than if y 261 /dev/bcache<N> This will work faster than if you dmcrypt both the backing
269 and caching devices and then install bcache on 262 and caching devices and then install bcache on top. [benchmarks?]
270 263
271 264
272 H) Stop/free a registered bcache to wipe and/o 265 H) Stop/free a registered bcache to wipe and/or recreate it
273 266
274 Suppose that you need to free up all bcache re 267 Suppose that you need to free up all bcache references so that you can
275 fdisk run and re-register a changed partition 268 fdisk run and re-register a changed partition table, which won't work
276 if there are any active backing or caching dev 269 if there are any active backing or caching devices left on it:
277 270
278 1) Is it present in /dev/bcache* ? (there are 271 1) Is it present in /dev/bcache* ? (there are times where it won't be)
279 272
280 If so, it's easy:: 273 If so, it's easy::
281 274
282 host:/sys/block/bcache0/bcache# echo 1 275 host:/sys/block/bcache0/bcache# echo 1 > stop
283 276
284 2) But if your backing device is gone, this wo 277 2) But if your backing device is gone, this won't work::
285 278
286 host:/sys/block/bcache0# cd bcache 279 host:/sys/block/bcache0# cd bcache
287 bash: cd: bcache: No such file or dire 280 bash: cd: bcache: No such file or directory
288 281
289 In this case, you may have to unregister th 282 In this case, you may have to unregister the dmcrypt block device that
290 references this bcache to free it up:: 283 references this bcache to free it up::
291 284
292 host:~# dmsetup remove oldds1 285 host:~# dmsetup remove oldds1
293 bcache: bcache_device_free() bcache0 s 286 bcache: bcache_device_free() bcache0 stopped
294 bcache: cache_set_free() Cache set 5bc 287 bcache: cache_set_free() Cache set 5bc072a8-ab17-446d-9744-e247949913c1 unregistered
295 288
296 This causes the backing bcache to be remove 289 This causes the backing bcache to be removed from /sys/fs/bcache and
297 then it can be reused. This would be true 290 then it can be reused. This would be true of any block device stacking
298 where bcache is a lower device. 291 where bcache is a lower device.
299 292
300 3) In other cases, you can also look in /sys/f 293 3) In other cases, you can also look in /sys/fs/bcache/::
301 294
302 host:/sys/fs/bcache# ls -l */{cache?,b 295 host:/sys/fs/bcache# ls -l */{cache?,bdev?}
303 lrwxrwxrwx 1 root root 0 Mar 5 09:39 296 lrwxrwxrwx 1 root root 0 Mar 5 09:39 0226553a-37cf-41d5-b3ce-8b1e944543a8/bdev1 -> ../../../devices/virtual/block/dm-1/bcache/
304 lrwxrwxrwx 1 root root 0 Mar 5 09:39 297 lrwxrwxrwx 1 root root 0 Mar 5 09:39 0226553a-37cf-41d5-b3ce-8b1e944543a8/cache0 -> ../../../devices/virtual/block/dm-4/bcache/
305 lrwxrwxrwx 1 root root 0 Mar 5 09:39 298 lrwxrwxrwx 1 root root 0 Mar 5 09:39 5bc072a8-ab17-446d-9744-e247949913c1/cache0 -> ../../../devices/pci0000:00/0000:00:01.0/0000:01:00.0/ata10/host9/target9:0:0/9:0:0:0/block/sdl/sdl2/bcache/
306 299
307 The device names will show which UUID is re 300 The device names will show which UUID is relevant, cd in that directory
308 and stop the cache:: 301 and stop the cache::
309 302
310 host:/sys/fs/bcache/5bc072a8-ab17-446d 303 host:/sys/fs/bcache/5bc072a8-ab17-446d-9744-e247949913c1# echo 1 > stop
311 304
312 This will free up bcache references and let 305 This will free up bcache references and let you reuse the partition for
313 other purposes. 306 other purposes.
314 307
315 308
316 309
317 Troubleshooting performance 310 Troubleshooting performance
318 --------------------------- 311 ---------------------------
319 312
320 Bcache has a bunch of config options and tunab 313 Bcache has a bunch of config options and tunables. The defaults are intended to
321 be reasonable for typical desktop and server w 314 be reasonable for typical desktop and server workloads, but they're not what you
322 want for getting the best possible numbers whe 315 want for getting the best possible numbers when benchmarking.
323 316
324 - Backing device alignment 317 - Backing device alignment
325 318
326 The default metadata size in bcache is 8k. 319 The default metadata size in bcache is 8k. If your backing device is
327 RAID based, then be sure to align this by a 320 RAID based, then be sure to align this by a multiple of your stride
328 width using `bcache make --data-offset`. If !! 321 width using `make-bcache --data-offset`. If you intend to expand your
329 disk array in the future, then multiply a s 322 disk array in the future, then multiply a series of primes by your
330 raid stripe size to get the disk multiples 323 raid stripe size to get the disk multiples that you would like.
331 324
332 For example: If you have a 64k stripe size 325 For example: If you have a 64k stripe size, then the following offset
333 would provide alignment for many common RAI 326 would provide alignment for many common RAID5 data spindle counts::
334 327
335 64k * 2*2*2*3*3*5*7 bytes = 161280k 328 64k * 2*2*2*3*3*5*7 bytes = 161280k
336 329
337 That space is wasted, but for only 157.5MB 330 That space is wasted, but for only 157.5MB you can grow your RAID 5
338 volume to the following data-spindle counts 331 volume to the following data-spindle counts without re-aligning::
339 332
340 3,4,5,6,7,8,9,10,12,14,15,18,20,21 ... 333 3,4,5,6,7,8,9,10,12,14,15,18,20,21 ...
341 334
342 - Bad write performance 335 - Bad write performance
343 336
344 If write performance is not what you expect 337 If write performance is not what you expected, you probably wanted to be
345 running in writeback mode, which isn't the 338 running in writeback mode, which isn't the default (not due to a lack of
346 maturity, but simply because in writeback m 339 maturity, but simply because in writeback mode you'll lose data if something
347 happens to your SSD):: 340 happens to your SSD)::
348 341
349 # echo writeback > /sys/block/bcache0/ 342 # echo writeback > /sys/block/bcache0/bcache/cache_mode
350 343
351 - Bad performance, or traffic not going to th 344 - Bad performance, or traffic not going to the SSD that you'd expect
352 345
353 By default, bcache doesn't cache everything 346 By default, bcache doesn't cache everything. It tries to skip sequential IO -
354 because you really want to be caching the r 347 because you really want to be caching the random IO, and if you copy a 10
355 gigabyte file you probably don't want that 348 gigabyte file you probably don't want that pushing 10 gigabytes of randomly
356 accessed data out of your cache. 349 accessed data out of your cache.
357 350
358 But if you want to benchmark reads from cac 351 But if you want to benchmark reads from cache, and you start out with fio
359 writing an 8 gigabyte test file - so you wa 352 writing an 8 gigabyte test file - so you want to disable that::
360 353
361 # echo 0 > /sys/block/bcache0/bcache/s 354 # echo 0 > /sys/block/bcache0/bcache/sequential_cutoff
362 355
363 To set it back to the default (4 mb), do:: 356 To set it back to the default (4 mb), do::
364 357
365 # echo 4M > /sys/block/bcache0/bcache/ 358 # echo 4M > /sys/block/bcache0/bcache/sequential_cutoff
366 359
367 - Traffic's still going to the spindle/still 360 - Traffic's still going to the spindle/still getting cache misses
368 361
369 In the real world, SSDs don't always keep u 362 In the real world, SSDs don't always keep up with disks - particularly with
370 slower SSDs, many disks being cached by one 363 slower SSDs, many disks being cached by one SSD, or mostly sequential IO. So
371 you want to avoid being bottlenecked by the 364 you want to avoid being bottlenecked by the SSD and having it slow everything
372 down. 365 down.
373 366
374 To avoid that bcache tracks latency to the 367 To avoid that bcache tracks latency to the cache device, and gradually
375 throttles traffic if the latency exceeds a 368 throttles traffic if the latency exceeds a threshold (it does this by
376 cranking down the sequential bypass). 369 cranking down the sequential bypass).
377 370
378 You can disable this if you need to by sett 371 You can disable this if you need to by setting the thresholds to 0::
379 372
380 # echo 0 > /sys/fs/bcache/<cache set>/ 373 # echo 0 > /sys/fs/bcache/<cache set>/congested_read_threshold_us
381 # echo 0 > /sys/fs/bcache/<cache set>/ 374 # echo 0 > /sys/fs/bcache/<cache set>/congested_write_threshold_us
382 375
383 The default is 2000 us (2 milliseconds) for 376 The default is 2000 us (2 milliseconds) for reads, and 20000 for writes.
384 377
385 - Still getting cache misses, of the same dat 378 - Still getting cache misses, of the same data
386 379
387 One last issue that sometimes trips people 380 One last issue that sometimes trips people up is actually an old bug, due to
388 the way cache coherency is handled for cach 381 the way cache coherency is handled for cache misses. If a btree node is full,
389 a cache miss won't be able to insert a key 382 a cache miss won't be able to insert a key for the new data and the data
390 won't be written to the cache. 383 won't be written to the cache.
391 384
392 In practice this isn't an issue because as 385 In practice this isn't an issue because as soon as a write comes along it'll
393 cause the btree node to be split, and you n 386 cause the btree node to be split, and you need almost no write traffic for
394 this to not show up enough to be noticeable 387 this to not show up enough to be noticeable (especially since bcache's btree
395 nodes are huge and index large regions of t 388 nodes are huge and index large regions of the device). But when you're
396 benchmarking, if you're trying to warm the 389 benchmarking, if you're trying to warm the cache by reading a bunch of data
397 and there's no other traffic - that can be 390 and there's no other traffic - that can be a problem.
398 391
399 Solution: warm the cache by doing writes, o 392 Solution: warm the cache by doing writes, or use the testing branch (there's
400 a fix for the issue there). 393 a fix for the issue there).
401 394
402 395
403 Sysfs - backing device 396 Sysfs - backing device
404 ---------------------- 397 ----------------------
405 398
406 Available at /sys/block/<bdev>/bcache, /sys/bl 399 Available at /sys/block/<bdev>/bcache, /sys/block/bcache*/bcache and
407 (if attached) /sys/fs/bcache/<cset-uuid>/bdev* 400 (if attached) /sys/fs/bcache/<cset-uuid>/bdev*
408 401
409 attach 402 attach
410 Echo the UUID of a cache set to this file to 403 Echo the UUID of a cache set to this file to enable caching.
411 404
412 cache_mode 405 cache_mode
413 Can be one of either writethrough, writeback 406 Can be one of either writethrough, writeback, writearound or none.
414 407
415 clear_stats 408 clear_stats
416 Writing to this file resets the running tota 409 Writing to this file resets the running total stats (not the day/hour/5 minute
417 decaying versions). 410 decaying versions).
418 411
419 detach 412 detach
420 Write to this file to detach from a cache se 413 Write to this file to detach from a cache set. If there is dirty data in the
421 cache, it will be flushed first. 414 cache, it will be flushed first.
422 415
423 dirty_data 416 dirty_data
424 Amount of dirty data for this backing device 417 Amount of dirty data for this backing device in the cache. Continuously
425 updated unlike the cache set's version, but 418 updated unlike the cache set's version, but may be slightly off.
426 419
427 label 420 label
428 Name of underlying device. 421 Name of underlying device.
429 422
430 readahead 423 readahead
431 Size of readahead that should be performed. 424 Size of readahead that should be performed. Defaults to 0. If set to e.g.
432 1M, it will round cache miss reads up to tha 425 1M, it will round cache miss reads up to that size, but without overlapping
433 existing cache entries. 426 existing cache entries.
434 427
435 running 428 running
436 1 if bcache is running (i.e. whether the /de 429 1 if bcache is running (i.e. whether the /dev/bcache device exists, whether
437 it's in passthrough mode or caching). 430 it's in passthrough mode or caching).
438 431
439 sequential_cutoff 432 sequential_cutoff
440 A sequential IO will bypass the cache once i 433 A sequential IO will bypass the cache once it passes this threshold; the
441 most recent 128 IOs are tracked so sequentia 434 most recent 128 IOs are tracked so sequential IO can be detected even when
442 it isn't all done at once. 435 it isn't all done at once.
443 436
444 sequential_merge 437 sequential_merge
445 If non zero, bcache keeps a list of the last 438 If non zero, bcache keeps a list of the last 128 requests submitted to compare
446 against all new requests to determine which 439 against all new requests to determine which new requests are sequential
447 continuations of previous requests for the p 440 continuations of previous requests for the purpose of determining sequential
448 cutoff. This is necessary if the sequential 441 cutoff. This is necessary if the sequential cutoff value is greater than the
449 maximum acceptable sequential size for any s 442 maximum acceptable sequential size for any single request.
450 443
451 state 444 state
452 The backing device can be in one of four dif 445 The backing device can be in one of four different states:
453 446
454 no cache: Has never been attached to a cache 447 no cache: Has never been attached to a cache set.
455 448
456 clean: Part of a cache set, and there is no 449 clean: Part of a cache set, and there is no cached dirty data.
457 450
458 dirty: Part of a cache set, and there is cac 451 dirty: Part of a cache set, and there is cached dirty data.
459 452
460 inconsistent: The backing device was forcibl 453 inconsistent: The backing device was forcibly run by the user when there was
461 dirty data cached but the cache set was unav 454 dirty data cached but the cache set was unavailable; whatever data was on the
462 backing device has likely been corrupted. 455 backing device has likely been corrupted.
463 456
464 stop 457 stop
465 Write to this file to shut down the bcache d 458 Write to this file to shut down the bcache device and close the backing
466 device. 459 device.
467 460
468 writeback_delay 461 writeback_delay
469 When dirty data is written to the cache and 462 When dirty data is written to the cache and it previously did not contain
470 any, waits some number of seconds before ini 463 any, waits some number of seconds before initiating writeback. Defaults to
471 30. 464 30.
472 465
473 writeback_percent 466 writeback_percent
474 If nonzero, bcache tries to keep around this 467 If nonzero, bcache tries to keep around this percentage of the cache dirty by
475 throttling background writeback and using a 468 throttling background writeback and using a PD controller to smoothly adjust
476 the rate. 469 the rate.
477 470
478 writeback_rate 471 writeback_rate
479 Rate in sectors per second - if writeback_pe 472 Rate in sectors per second - if writeback_percent is nonzero, background
480 writeback is throttled to this rate. Continu 473 writeback is throttled to this rate. Continuously adjusted by bcache but may
481 also be set by the user. 474 also be set by the user.
482 475
483 writeback_running 476 writeback_running
484 If off, writeback of dirty data will not tak 477 If off, writeback of dirty data will not take place at all. Dirty data will
485 still be added to the cache until it is most 478 still be added to the cache until it is mostly full; only meant for
486 benchmarking. Defaults to on. 479 benchmarking. Defaults to on.
487 480
488 Sysfs - backing device stats 481 Sysfs - backing device stats
489 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 482 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~
490 483
491 There are directories with these numbers for a 484 There are directories with these numbers for a running total, as well as
492 versions that decay over the past day, hour an 485 versions that decay over the past day, hour and 5 minutes; they're also
493 aggregated in the cache set directory as well. 486 aggregated in the cache set directory as well.
494 487
495 bypassed 488 bypassed
496 Amount of IO (both reads and writes) that ha 489 Amount of IO (both reads and writes) that has bypassed the cache
497 490
498 cache_hits, cache_misses, cache_hit_ratio 491 cache_hits, cache_misses, cache_hit_ratio
499 Hits and misses are counted per individual I 492 Hits and misses are counted per individual IO as bcache sees them; a
500 partial hit is counted as a miss. 493 partial hit is counted as a miss.
501 494
502 cache_bypass_hits, cache_bypass_misses 495 cache_bypass_hits, cache_bypass_misses
503 Hits and misses for IO that is intended to s 496 Hits and misses for IO that is intended to skip the cache are still counted,
504 but broken out here. 497 but broken out here.
505 498
506 cache_miss_collisions 499 cache_miss_collisions
507 Counts instances where data was going to be 500 Counts instances where data was going to be inserted into the cache from a
508 cache miss, but raced with a write and data 501 cache miss, but raced with a write and data was already present (usually 0
509 since the synchronization for cache misses w 502 since the synchronization for cache misses was rewritten)
>> 503
>> 504 cache_readaheads
>> 505 Count of times readahead occurred.
510 506
511 Sysfs - cache set 507 Sysfs - cache set
512 ~~~~~~~~~~~~~~~~~ 508 ~~~~~~~~~~~~~~~~~
513 509
514 Available at /sys/fs/bcache/<cset-uuid> 510 Available at /sys/fs/bcache/<cset-uuid>
515 511
516 average_key_size 512 average_key_size
517 Average data per key in the btree. 513 Average data per key in the btree.
518 514
519 bdev<0..n> 515 bdev<0..n>
520 Symlink to each of the attached backing devi 516 Symlink to each of the attached backing devices.
521 517
522 block_size 518 block_size
523 Block size of the cache devices. 519 Block size of the cache devices.
524 520
525 btree_cache_size 521 btree_cache_size
526 Amount of memory currently used by the btree 522 Amount of memory currently used by the btree cache
527 523
528 bucket_size 524 bucket_size
529 Size of buckets 525 Size of buckets
530 526
531 cache<0..n> 527 cache<0..n>
532 Symlink to each of the cache devices compris 528 Symlink to each of the cache devices comprising this cache set.
533 529
534 cache_available_percent 530 cache_available_percent
535 Percentage of cache device which doesn't con 531 Percentage of cache device which doesn't contain dirty data, and could
536 potentially be used for writeback. This doe 532 potentially be used for writeback. This doesn't mean this space isn't used
537 for clean cached data; the unused statistic 533 for clean cached data; the unused statistic (in priority_stats) is typically
538 much lower. 534 much lower.
539 535
540 clear_stats 536 clear_stats
541 Clears the statistics associated with this c 537 Clears the statistics associated with this cache
542 538
543 dirty_data 539 dirty_data
544 Amount of dirty data is in the cache (update 540 Amount of dirty data is in the cache (updated when garbage collection runs).
545 541
546 flash_vol_create 542 flash_vol_create
547 Echoing a size to this file (in human readab 543 Echoing a size to this file (in human readable units, k/M/G) creates a thinly
548 provisioned volume backed by the cache set. 544 provisioned volume backed by the cache set.
549 545
550 io_error_halflife, io_error_limit 546 io_error_halflife, io_error_limit
551 These determines how many errors we accept b 547 These determines how many errors we accept before disabling the cache.
552 Each error is decayed by the half life (in # 548 Each error is decayed by the half life (in # ios). If the decaying count
553 reaches io_error_limit dirty data is written 549 reaches io_error_limit dirty data is written out and the cache is disabled.
554 550
555 journal_delay_ms 551 journal_delay_ms
556 Journal writes will delay for up to this man 552 Journal writes will delay for up to this many milliseconds, unless a cache
557 flush happens sooner. Defaults to 100. 553 flush happens sooner. Defaults to 100.
558 554
559 root_usage_percent 555 root_usage_percent
560 Percentage of the root btree node in use. I 556 Percentage of the root btree node in use. If this gets too high the node
561 will split, increasing the tree depth. 557 will split, increasing the tree depth.
562 558
563 stop 559 stop
564 Write to this file to shut down the cache se 560 Write to this file to shut down the cache set - waits until all attached
565 backing devices have been shut down. 561 backing devices have been shut down.
566 562
567 tree_depth 563 tree_depth
568 Depth of the btree (A single node btree has 564 Depth of the btree (A single node btree has depth 0).
569 565
570 unregister 566 unregister
571 Detaches all backing devices and closes the 567 Detaches all backing devices and closes the cache devices; if dirty data is
572 present it will disable writeback caching an 568 present it will disable writeback caching and wait for it to be flushed.
573 569
574 Sysfs - cache set internal 570 Sysfs - cache set internal
575 ~~~~~~~~~~~~~~~~~~~~~~~~~~ 571 ~~~~~~~~~~~~~~~~~~~~~~~~~~
576 572
577 This directory also exposes timings for a numb 573 This directory also exposes timings for a number of internal operations, with
578 separate files for average duration, average f 574 separate files for average duration, average frequency, last occurrence and max
579 duration: garbage collection, btree read, btre 575 duration: garbage collection, btree read, btree node sorts and btree splits.
580 576
581 active_journal_entries 577 active_journal_entries
582 Number of journal entries that are newer tha 578 Number of journal entries that are newer than the index.
583 579
584 btree_nodes 580 btree_nodes
585 Total nodes in the btree. 581 Total nodes in the btree.
586 582
587 btree_used_percent 583 btree_used_percent
588 Average fraction of btree in use. 584 Average fraction of btree in use.
589 585
590 bset_tree_stats 586 bset_tree_stats
591 Statistics about the auxiliary search trees 587 Statistics about the auxiliary search trees
592 588
593 btree_cache_max_chain 589 btree_cache_max_chain
594 Longest chain in the btree node cache's hash 590 Longest chain in the btree node cache's hash table
595 591
596 cache_read_races 592 cache_read_races
597 Counts instances where while data was being 593 Counts instances where while data was being read from the cache, the bucket
598 was reused and invalidated - i.e. where the 594 was reused and invalidated - i.e. where the pointer was stale after the read
599 completed. When this occurs the data is rere 595 completed. When this occurs the data is reread from the backing device.
600 596
601 trigger_gc 597 trigger_gc
602 Writing to this file forces garbage collecti 598 Writing to this file forces garbage collection to run.
603 599
604 Sysfs - Cache device 600 Sysfs - Cache device
605 ~~~~~~~~~~~~~~~~~~~~ 601 ~~~~~~~~~~~~~~~~~~~~
606 602
607 Available at /sys/block/<cdev>/bcache 603 Available at /sys/block/<cdev>/bcache
608 604
609 block_size 605 block_size
610 Minimum granularity of writes - should match 606 Minimum granularity of writes - should match hardware sector size.
611 607
612 btree_written 608 btree_written
613 Sum of all btree writes, in (kilo/mega/giga) 609 Sum of all btree writes, in (kilo/mega/giga) bytes
614 610
615 bucket_size 611 bucket_size
616 Size of buckets 612 Size of buckets
617 613
618 cache_replacement_policy 614 cache_replacement_policy
619 One of either lru, fifo or random. 615 One of either lru, fifo or random.
620 616
621 discard 617 discard
622 Boolean; if on a discard/TRIM will be issued 618 Boolean; if on a discard/TRIM will be issued to each bucket before it is
623 reused. Defaults to off, since SATA TRIM is 619 reused. Defaults to off, since SATA TRIM is an unqueued command (and thus
624 slow). 620 slow).
625 621
626 freelist_percent 622 freelist_percent
627 Size of the freelist as a percentage of nbuc 623 Size of the freelist as a percentage of nbuckets. Can be written to to
628 increase the number of buckets kept on the f 624 increase the number of buckets kept on the freelist, which lets you
629 artificially reduce the size of the cache at 625 artificially reduce the size of the cache at runtime. Mostly for testing
630 purposes (i.e. testing how different size ca 626 purposes (i.e. testing how different size caches affect your hit rate), but
631 since buckets are discarded when they move o 627 since buckets are discarded when they move on to the freelist will also make
632 the SSD's garbage collection easier by effec 628 the SSD's garbage collection easier by effectively giving it more reserved
633 space. 629 space.
634 630
635 io_errors 631 io_errors
636 Number of errors that have occurred, decayed 632 Number of errors that have occurred, decayed by io_error_halflife.
637 633
638 metadata_written 634 metadata_written
639 Sum of all non data writes (btree writes and 635 Sum of all non data writes (btree writes and all other metadata).
640 636
641 nbuckets 637 nbuckets
642 Total buckets in this cache 638 Total buckets in this cache
643 639
644 priority_stats 640 priority_stats
645 Statistics about how recently data in the ca 641 Statistics about how recently data in the cache has been accessed.
646 This can reveal your working set size. Unus 642 This can reveal your working set size. Unused is the percentage of
647 the cache that doesn't contain any data. Me 643 the cache that doesn't contain any data. Metadata is bcache's
648 metadata overhead. Average is the average p 644 metadata overhead. Average is the average priority of cache buckets.
649 Next is a list of quantiles with the priorit 645 Next is a list of quantiles with the priority threshold of each.
650 646
651 written 647 written
652 Sum of all data that has been written to the 648 Sum of all data that has been written to the cache; comparison with
653 btree_written gives the amount of write infl 649 btree_written gives the amount of write inflation in bcache.
Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.