Newer
Older
# What's DiskSim?
[DiskSim](http://www.pdl.cmu.edu/DiskSim/index.shtml) is an efficient, accurate, highly-configurable disk system simulator originally developed at the University of Michigan and enhanced at CMU to support research into various aspects of storage subsystem architecture. It is written in C and requires no special system software (just basic POSIX interfaces). DiskSim includes modules for most secondary storage components of interest, including device drivers, buses, controllers, adapters, and disk drives. DiskSim also includes support for a number of externally-provided trace formats and internally-generated synthetic workloads, and includes hooks for inclusion in a larger scale system-level simulator.
# Why to use DiskSim?
Because it's unprecedented accuracy:
# DiskSim builder and scripts
This tool helps with building DiskSim v4.0 in these ways:
* base sources or these with [DIXtrac] ) tool were downloaded from: http://www.pdl.cmu.edu/Dixtrac/index.shtml
* patch sources by Microsoft [SSD extension] were downloaded from: https://www.microsoft.com/en-us/download/details.aspx?id=52332
* patch sources by Oliver Liu [64bit patch] were downloaded from: https://github.com/myidpt/PFSsim/tree/master/disksim/64bit-ssd-patch-files
* enforce C standard and optimization, so `disksim` can be compiled by f.e. `gcc` v5.4.0 and can run faster
* run validation tests coming with sources
Folder `scripts/` holds other useful tools including a generator of **JBOD|RAID{0|1|3|4|5}** _parv_ files.
### Clone this repo
```
$ git clone git@gitlab.fit.cvut.cz:berusjan/disksim-4.0.git
$ cd disksim-4.0
```
### Install requirements
It requires installed C development tools, `bison` (YACC-compatible parser generator) and `flex` (fast lexical analyzer generator).
The example of installation command for Ubuntu distribution:
```
sudo apt-get install build-essential bison flex
```
### Help for make
```
$ make help
```
```
Make patched DiskSim
Usage:
compile ..... compile disksim in build/disksim-4.0/
install ..... install binaries to bin/
test ........ run valadation tests
help ........ print this help screen
```
### Build DiskSim
```
$ make
```
```
[DISKSIM] Compile ...
chmod +x 'build/disksim-4.0/libparam/'*.pl
make -C 'build/disksim-4.0'
make[1]: Vstupuje se do adresáře „/home/oem/BP/My_BP/patches_all/patched/build/disksim-4.0“
make -C libddbg
make[2]: Vstupuje se do adresáře „/home/oem/BP/My_BP/patches_all/patched/build/disksim-4.0/libddbg“
cc -O2 -std=gnu90 -g -I../libparam -c -o ddbg_assert.o ddbg_assert.c
cc -O2 -std=gnu90 -g -I../libparam -c -o ddbg_trace.o ddbg_trace.c
ar cru libddbg.a ddbg_assert.o ddbg_trace.o
ar: `u' modifier ignored since `D' is the default (see `U')
mkdir -p include/libddbg lib
cp -p libddbg.h include/libddbg
cp -p libddbg.a lib
make[2]: Opouští se adresář „/home/oem/BP/My_BP/patches_all/patched/build/disksim-4.0/libddbg“
...
OK, installed binaries:
-rwxrwxr-x 1 jake jake 606424 pro 7 03:18 bin/disksim
-rwxrwxr-x 1 jake jake 51784 pro 7 03:18 bin/dxcachectl
-rwxrwxr-x 1 jake jake 72536 pro 7 03:18 bin/dx_layout
-rwxrwxr-x 1 jake jake 197560 pro 7 03:18 bin/dx_mech
-rwxrwxr-x 1 jake jake 55960 pro 7 03:18 bin/dxreplay
-rwxrwxr-x 1 jake jake 700120 pro 7 03:18 bin/dx_rest
-rwxrwxr-x 1 jake jake 197784 pro 7 03:18 bin/dx_seeks
-rwxrwxr-x 1 jake jake 172888 pro 7 03:18 bin/dx_skew_curve
-rwxrwxr-x 1 jake jake 168696 pro 7 03:18 bin/dx_skews_simple
-rwxrwxr-x 1 jake jake 72376 pro 7 03:18 bin/dx_stat
-rwxrwxr-x 1 jake jake 10528 pro 7 03:18 bin/hplcomb
-rwxrwxr-x 1 jake jake 22792 pro 7 03:18 bin/mems_seektest
-rwxrwxr-x 1 jake jake 610552 pro 7 03:18 bin/physim
-rwxrwxr-x 1 jake jake 10480 pro 7 03:18 bin/rms
[DISKSIM] Run validation test build/disksim-4.0/valid/runvalid ...
cd 'build/disksim-4.0/valid/' && sh 'runvalid'
These results represent actual drive validation experiments
...
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
```
### Run validation tests
```
$ make test
```
```
[DISKSIM] Run validation test build/disksim-4.0/valid/runvalid ...
cd 'build/disksim-4.0/valid/' && sh 'runvalid'
These results represent actual drive validation experiments
QUANTUM_QM39100TD-SW (rms should be about 0.378)
rms = 0.378078
SEAGATE_ST32171W (rms should be about 0.349)
rms = 0.347863
...
```
## Scripts
### Go to script folder
```
$ cd script
```
### `disksim.sh`
```
$ ./disksim.sh
```
```
DiskSim wrapper
Usage:
./disksim.sh [-q] <param_file> [ <output_file> [ <trace_file> [ <trace_type> [ <par_override> ... ] ] ] ]
Options:
-q ............... Quiet - supress verbose messages
Arguments:
<param_file> ..... Name of the parameter file
Set - for standard input
<output_file> .... Name of the output file
Output can be directed to stdout by specifying 'stdout' as a basename,
and it quiet verbose messages. That's a default
<trace_file> ..... Trace file to be used as input
Input is taken from stdin when 'stdin' is specified,
of course not together with <param_file> from 'stdin'
Synthetic workload generation portion of the simulator is enabled
instead of a trace file if it is empty, '-' or '0'
<trace_type> ..... Format of the trace input, like 'ascii', 'validate', 'raw', 'hpl' ...
Set to 'ascii' if it's empty, '-' or '0'
<par_override> ... Replacement for default parameter values or parameter values from <param_file>
Notes:
* disksim runs in a folder with diskspecs (diskspecs/) and temporarily copies <param_file> here
* it prints colored messages and resulted file list
```
### `gen_raid.sh`
```
$ ./gen_raid.sh -h
```
```
Usage:
./gen_raid.sh [options] jbod|raid{0|1|3|4|5} [<number_of_disks>]
./gen_raid.sh -h
Arguments:
jbod|raid{0|1|3|4|5} raid type
<number_of_disks> integer number of disks (by default 3)
Options:
-h show this help screen and exit
-v increase output verbosity
-m print mapping requests
-o FILE output parv filename; - for stdout (by default -)
-r INT number of I/O requests to generate (by default 1000)
-S FLOAT probability of sequential access <0.0,1.0> (by default 0.5)
-R FLOAT probability of read access <0.0,1.0> (by default 0.5)
-p INT number of processes to genereate load (by default 1)
-u INT stripe unit (by default 64)
-s REQUEST_SIZE disksim style request size,
default: 'exponential, 0.0, 8.0'
types of sizes:
uniform - requiring two floats - minimum and maximum
e.g. -s 'uniform, 1.0, 8.0'
normal - requiring two floats - mean and variance values
- (stredni hodnota a hodnota rozptylu)
- second value must be nonnegative
e.g. -s 'normal, 5.0, 4,0'
exponential - requiring two floats - base and mean values
e.g. -s 'exponential, 0.0, 16.0'
-d DISK_NAME disksim disk to simulate (by default QUANTUM_TORNADO_validate)
available rotational disks:
DEC_RZ26_validate
HP_C2247A_validate
HP_C2249A
HP_C2490A_validate
HP_C3323A
HP_C3323A_validate
IBM_DNES-309170W_validate
QUANTUM_QM39100TD-SW
QUANTUM_TORNADO_validate
SEAGATE_ST32171W
SEAGATE_ST34501N_validate
SEAGATE_ST39102LW_validate
Seagate_ST41601N_validate
available mem disks:
MEMS_G1
MEMS_G2
MEMS_G3
available ssd disks:
SSD_IOZONE
```
### `run_raid_procs.sh`
```
$ ./run_raid_procs.sh
```
```
Collect and represent disksim metrics
Usage for collecting metrics:
./run_raid_procs.sh -p <max_num_of_processes> [gen_raid_options] jbod|raid{0|1|3|4|5} [<number_of_disks>]
Usage for get ASCII metrics:
./run_raid_procs.sh --ascii <metric> ...
Usage for get gnuplot metrics:
./run_raid_procs.sh ---gnuplot <metric> ...
Availaible metrics:
'Average # requests'
'Average queue length'
'Avg # read requests'
'Avg # write requests'
'Base SPTF/SDF Different'
'Batch size average'
'Batch size maximum'
'Batch size std.dev.'
'Completely idle time'
'Critical Read Response time average'
'Critical Read Response time maximum'
'Critical Read Response time std.dev.'
'Critical Reads'
'Critical Write Response time average'
'Critical Write Response time maximum'
'Critical Write Response time std.dev.'
'Critical Writes'
'End # requests'
'End queued requests'
'Idle period length average'
'Idle period length maximum'
'Idle period length std.dev.'
'Instantaneous queue length average'
'Instantaneous queue length maximum'
'Instantaneous queue length std.dev.'
'Inter-arrival time average'
'Inter-arrival time maximum'
'Inter-arrival time std.dev.'
'Max # read requests'
'Max # write requests'
'Maximum # requests'
'Maximum queue length'
'Non-Critical Read Response time average'
'Non-Critical Read Response time maximum'
'Non-Critical Read Response time std.dev.'
'Non-Critical Reads'
'Non-Critical Write Response time average'
'Non-Critical Write Response time maximum'
'Non-Critical Write Response time std.dev.'
'Non-Critical Writes'
'Number of batches'
'Number of idle periods'
'Number of reads'
'Number of writes'
'Overlaps combined'
'Physical access time average'
'Physical access time maximum'
'Physical access time std.dev.'
'Priority SPTF/SDF Different'
'Queue time average'
'Queue time maximum'
'Queue time std.dev.'
'Read inter-arrival average'
'Read inter-arrival maximum'
'Read inter-arrival std.dev.'
'Read overlaps combined'
'Read request size average'
'Read request size maximum'
'Read request size std.dev.'
'Request size average'
'Request size maximum'
'Request size std.dev.'
'Requests per second'
'Response time average'
'Response time maximum'
'Response time std.dev.'
'runlistlen'
'runoutstanding'
'Sequential reads'
'Sequential writes'
'setsize'
'simtime'
'Sub-optimal mapping penalty average'
'Sub-optimal mapping penalty maximum'
'Sub-optimal mapping penalty std.dev.'
'Timeout SPTF/SDF Different'
'Total Requests handled'
'warmuptime'
'Write inter-arrival average'
'Write inter-arrival maximum'
'Write inter-arrival std.dev.'
'Write request size average'
'Write request size maximum'
'Write request size std.dev.'
```
### Examples
#### Print overall I/O stats from RAID5
```
$ ./gen_raid.sh raid5 | ./disksim.sh - | ./filter_overall.pl
```
```
Total Requests handled: 1000
Requests per second: 52.113620
Completely idle time: 75.004636 0.003909
Response time average: 7.294895
...
```
#### Print IOPS stats from RAID0
```
$ ./gen_raid.sh raid0 | ./disksim.sh - | ./filter_iops.pl
```
```
Average latency [ms]: 4.18640
RAID I/O per second: 62.18773
RAID kB per second: 392.03146
DISK#0 I/O per second: 24.31540
DISK#0 kB per second: 140.43385
```
#### Print mappings for all supported RAIDs
```
$ for raid in jbod raid0 raid1 raid3 raid4 raid5; do
./gen_raid.sh -m -r20 -R0.0 -u1 ${raid} 4 | ./disksim.sh - | ./filter_map_requests.pl
done
```
```
Mapping for logorg jbod:
block disk0
---------+--------------------
1992757 0
1992758 1
1992759 2
1992760 3
1992761 4
1992762 5
1992763 6
1992764 7
1992765 8
1992766 9
1992767 10
1992768 11
1992769 12
1992770 13
1992771 14
1992772 15
1992773 16
1992774 17
1992775 18
1992776 19
Mapping for logorg raid0:
block disk0 disk1 disk2 disk3
---------+--------------------+--------------------+--------------------+--------------------
498189 0 1 2
498190 3 4 5 6
498191 7 8 9 10
498192 11 12 13 14
498193 15 16 17 18
498194 19
Mapping for logorg raid1:
block disk0 disk1 disk2 disk3
---------+--------------------+--------------------+--------------------+--------------------
1992757 0 0 0 0
1992758 1 1 1 1
1992759 2 2 2 2
1992760 3 3 3 3
1992761 4 4 4 4
1992762 5 5 5 5
1992763 6 6 6 6
1992764 7 7 7 7
1992765 8 8 8 8
1992766 9 9 9 9
1992767 10 10 10 10
1992768 11 11 11 11
1992769 12 12 12 12
1992770 13 13 13 13
1992771 14 14 14 14
1992772 15 15 15 15
1992773 16 16 16 16
1992774 17 17 17 17
1992775 18 18 18 18
1992776 19 19 19 19
Mapping for logorg raid3:
block disk0 disk1 disk2 disk3
---------+--------------------+--------------------+--------------------+--------------------
664252 0 1 0,1
664253 2 3 4 2,3,4
664254 5 6 7 5,6,7
664255 8 9 10 8,9,10
664256 11 12 13 11,12,13
664257 14 15 16 14,15,16
664258 17 18 19 17,18,19
Mapping for logorg raid4:
block disk0 disk1 disk2 disk3
---------+--------------------+--------------------+--------------------+--------------------
664252 0 1 0,1
664253 2 3 4 2,3,4
664254 5 6 7 5,6,7
664255 8 9 10 8,9,10
664256 11 12 13 11,12,13
664257 14 15 16 14,15,16
664258 17 18 19 17,18,19
Mapping for logorg raid5:
block disk0 disk1 disk2 disk3
---------+--------------------+--------------------+--------------------+--------------------
664252 0 1 0,1
664253 3 4 2,3,4 2
664254 7 5,6,7 5 6
664255 8,9,10 8 9 10
664256 11 12 13 11,12,13
664257 15 16 14,15,16 14
664258 19 17,18,19 17 18
```
#### Collect metrics for RAID5 with 10000 requests for 1-100 parallel processes
```
$ ./run_raid_procs.sh -p100 -r10000 raid5
```
```
Collecting metrics for 100 threads ... done (in .work/metrics_*.out)
```
#### Print collected metric for 1-100 parallel processes to a screen
```
./run_raid_procs.sh --ascii 'Response time average'
```
```
00000001 threads: 7.430611
00000002 threads: 11.030188
00000003 threads: 13.360172
00000004 threads: 15.235009
00000005 threads: 16.508526
00000006 threads: 17.762317
...
```
#### Show collected metrics for 1-100 parallel processes as a pictures
```
./run_raid_procs.sh --gnuplot 'Average # requests' 'Response time average' 'Write inter-arrival average' 'Completely idle time'
```
