-
Notifications
You must be signed in to change notification settings - Fork 0
/
index.html
368 lines (310 loc) · 16.3 KB
/
index.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
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
<!DOCTYPE html>
<html>
<head>
<meta charset="utf-8">
<title>Passive testing of production systems based on model inference. | William Durand</title>
<link rel="stylesheet" href="reveal.js/css/reveal.min.css">
<link rel="stylesheet" href="reveal.js/css/theme/simple.css" id="theme">
<link rel="stylesheet" href="reveal.js/lib/css/zenburn.css">
<link rel="stylesheet" href="font-awesome/css/font-awesome.min.css">
<link rel="stylesheet" href="css/fix-fontawesome-reveal-js.css">
<link rel="stylesheet" href="css/custom.css">
<script>
if (window.location.search.match(/print-pdf/gi)) {
document.write('<link rel="stylesheet" href="css/pdf.css" type="text/css">');
}
</script>
<!--[if lt IE 9]>
<script src="reveal.js/lib/js/html5shiv.js"></script>
<![endif]-->
</head>
<body>
<div class="reveal">
<div class="slides">
<section>
<h1 class="title">Passive testing of<br>production systems based<br>on model inference.</h1>
<em class="subtitle">
<u>William Durand</u>, Sébastien Salva
<br>
</em>
<em class="date">
September 22, 2015 <small><i class="fa fa-birthday-cake"></i></small> / MEMOCODE'15
</em>
<p class="sponsors">
<img src="images/blaisepascal.png" class="no-border" />
<img src="images/limos.png" class="no-border" />
<img src="images/michelin.jpg" class="no-border" />
</p>
</section>
<section data-background="images/chaine-des-puys.jpg"></section>
<section>
<section data-markdown>
<script type="text/template">
## Quick Tour @ Michelin
</script>
</section>
<section data-markdown>
<script type="text/template">
A **factory** is divided into several **workshops**,
<br>
one for each step of the manufacturing process.
</script>
</section>
<section data-markdown>
<script type="text/template">
A **production system** is composed of devices, production
machines, and one or more software to control them.
<br>
<i class="fa fa-info-circle"></i> In our case, we target a single workshop only.
</script>
</section>
<section data-markdown>
<script type="text/template">
Software exchange _information_ with points and machines
by sending and receiving **production messages**.
<br>
```
17-Sep-2015 23:29:59.50|17011|MSG_IN [pid: 1] [nsec: 8] [point: 1] ...
17-Sep-2015 23:29:59.61|17021|MSG_OUT [pid: 1] [nsec: 8] [point: 3] ...
17-Sep-2015 23:29:59.70|17011|MSG_IN [pid: 2] [nsec: 8] [point: 2] ...
```
<small>A simple example of 3 messages in a human readable format.</small>
</script>
</section>
<section data-markdown>
<script type="text/template">
**Production messages** are exchanged in a binary
format (custom protocols), through **centralized**
exchanging systems.
</script>
</section>
<section data-markdown>
<script type="text/template">
Each production message is tied to a **product** (e.g. tire),<br>identified
by a **product identifier** (`pid`).
<br>
Gathering all production messages related to a product<br>allows
to retrieve **what happened** to it (behaviours).
</script>
</section>
</section>
<section>
<section data-markdown>
<script type="text/template">
## Background
</script>
</section>
<section data-markdown>
<script type="text/template">
### Developement Teams POV
<br>
* 100+ applications running in **production**
* Not (fully) covered by tests
* Documentation most likely outdated
* MUST be maintained for ~20 years!
</script>
</section>
<section data-markdown>
<script type="text/template">
### Customers (Factories) POV
<br>
* Stability over anything else
* Maintenance periods are planned,<br>but rather long (> 1 week)
* 1h (unexpected) downtime = 50k $
</script>
</section>
<section data-markdown>
<script type="text/template">
**Testing** such production systems is **complex**,<br>
and takes a lot of time as it implies the physical<br>
devices, and there are **numerous behaviours**.
</script>
</section>
<section data-markdown>
<script type="text/template">
These behaviours could be formally described into a model.
<br>But writing such models would be **complicated** and **error prone**.
<br>
<i class="fa fa-info-circle"></i> Not suitable for Michelin applications.
</script>
</section>
<section data-markdown>
<script type="text/template">
### Our Approach (1/3)
<br>
By leveraging the information carried by the
messages, we<br>build **formal and exact
models** (STS) that describe functional behaviours
of a production System Under Analysis (`SUA`).
</script>
</section>
<section data-markdown>
<script type="text/template">
### Our Approach (2/3)
<br>
Michelin's exchanging systems **guarantee the order** in which<br>the production messages occured.
We **capture the messages directly into these systems** to avoid message loss, reordering, and/or
duplication of the production messages.
</script>
</section>
<section data-markdown>
<script type="text/template">
### Our Approach (3/3)
<br>
We take production messages from **another System Under Test** (`SUT`),
and we check whether `SUT` **conforms with** `SUA` (using two implementation
relations to define the notion of conformance).
</script>
</section>
</section>
<section>
<section data-markdown class="no-border">
<script type="text/template">
## The Big Picture
</script>
</section>
<section data-markdown class="no-border">
<script type="text/template">
### Model Inference
<br>
1. We **collect** production system traces (monitoring)
2. We **segment** these traces to create different complete<br>trace sets
(outlier detection approach)
3. We **build** (rather large) STS models from these sets
4. We **reduce** the models to obtain "usable" models
<br>
<p class="citation">
Durand, W., & Salva, S. (2015). Autofunk: An
Inference-Based Formal Model Generation Framework for
Production Systems. In FM 2015: Formal Methods (pp.
577-580). Springer International Publishing.
</p>
</script>
</section>
<section data-markdown class="no-border">
<script type="text/template">
### Model Reduction
<i class="fa fa-arrow-down"></i>
</script>
</section>
<section data-markdown class="no-border centered-table">
<script type="text/template">
### Model Inference Experimentation
<br>
10 million production messages (20 days)
<i class="fa fa-arrow-down"></i>
161,035 traces
<i class="fa fa-arrow-down"></i>
| `S` | `R(S)` |
|-----------------|----------------|
| 77,058 branches | 1,587 branches |
| 43,536 branches | 1,585 branches |
<br>
<small><i class="fa fa-info-circle"></i> It took 6 minutes to build the two models.</small>
</script>
</section>
<section>
<h3>In Depth Testing</h3>
<p>
<img src="images/overall_autofunk.png" class="no-border" />
</p>
</section>
<section data-markdown class="no-border">
<script type="text/template">
### Offline Passive Testing
<br>
* Two **implementation relations**:
- **Trace preorder** relation
- and our own **weaker** implementation relation
<br>
* Our testing algorithm relies on both to give verdicts
<br>
<i class="fa fa-warning"></i> Partial models = No `Fail` verdict
</script>
</section>
<section data-markdown class="no-border">
<script type="text/template">
### The Need for a Weaker Impl. Relation
<br>
_"Since I know that my model is not complete,
I am willing<br>to accept not standard behaviours till
a certain point."_
</script>
</section>
<section data-markdown class="no-border">
<script type="text/template">
### Experimentation
<br>
`SUA`: 53,996 traces
<br>
`SUT`: 25,047 traces
<br>
<i class="fa fa-arrow-down"></i>
**98%** are `Pass` traces. The remaining 2% are new<br>
behaviours that never occured before.
<br>
<small><i class="fa fa-info-circle"></i> It took 10 minutes to check conformance.</small>
</script>
</section>
<section data-markdown class="no-border">
<script type="text/template">
### Now, What?
<br>
**2%** still represents many traces, and can contain many false positive.
For Michelin engineers, it is still _"better than nothing"_.
<br>
* **Larger sets** of traces to build the models should
reduce the number of false positive
* But we should find a way to refine this **possibly fail** trace set
</script>
</section>
</section>
<section>
<section data-markdown>
<script type="text/template">
## Conclusion
<br>
* **Fast** passive testing framework for a specific context
* Model inference: the more production messages, the better!
* Testing: still too many possibly fail traces
</script>
</section>
</section>
<section data-markdown>
<script type="text/template">
## Future Work
<br>
* **Online** passive testing (just-in-time fault detection?)
* **Active testing** by leveraging the inferred models again
* Developing a way to focus on specific parts of the system
</script>
</section>
<section data-markdown>
<script type="text/template">
## Thank You.
### Questions?
</script>
</section>
</div>
</div>
<script type="text/javascript" src="reveal.js/lib/js/head.min.js"></script>
<script type="text/javascript" src="reveal.js/js/reveal.min.js"></script>
<script type="text/javascript">
Reveal.initialize({
slideNumber: !window.location.search.match(/print-pdf/gi),
history: true,
theme: Reveal.getQueryHash().theme,
transition: Reveal.getQueryHash().transition || 'concave',
dependencies: [
{ src: 'reveal.js/lib/js/classList.js', condition: function() { return !document.body.classList; } },
{ src: 'reveal.js/plugin/markdown/marked.js', condition: function() { return !!document.querySelector( '[data-markdown]' ); } },
{ src: 'reveal.js/plugin/markdown/markdown.js', condition: function() { return !!document.querySelector( '[data-markdown]' ); } },
{ src: 'js/highlight.js', async: true, callback: function() { hljs.initHighlightingOnLoad(); } },
]
});
</script>
</body>
</html>