Erlang The Driver behind WhatsApp s Success. Torben Hoffmann CTO, Erlang Solutions

Transcription

1 Erlang The Driver behind WhatsApp s Success Torben Hoffmann CTO, Erlang Solutions

2 Background

3 Background Erlanger since 2006

4 Background Erlanger since 2006 Happyness

5 Background Erlanger since 2006 Happyness Mission critical gateway for Tetra

6 Background Erlanger since 2006 Happyness Mission critical gateway for Tetra Hard work

7 Background Erlanger since 2006 Happyness Mission critical gateway for Tetra Hard work Major learnings

8 Background Erlanger since 2006 Happyness Mission critical gateway for Tetra Hard work Major learnings

9 Why this talk?

10 Why this talk? Explain the tech foundation for WhatsApp s succes

11 Why this talk? Explain the tech foundation for WhatsApp s succes Understand the design decisions that makes Erlang unique

12 Why this talk? Explain the tech foundation for WhatsApp s succes Understand the design decisions that makes Erlang unique Show how Erlang s features delivers business value

13 Why this talk? Explain the tech foundation for WhatsApp s succes Understand the design decisions that makes Erlang unique Show how Erlang s features delivers business value Spread the Erlang love

14 Religious Connection Er Lang Shen - Chinese God w/! a 3rd truth seeing eye Erlang Temple in Zunhua source: source:

15 Dealing with disbelievers Source: source: cb /villains/images/e/ef/saruman.jpg source:

16 Dealing with disbelievers Source: source: cb /villains/images/e/ef/saruman.jpg source:

17 Dealing with disbelievers Source: source: source: cb /villains/images/e/ef/saruman.jpg source:

18

19 Founded in 1999 soon after Erlang was released as open source

20 Founded in 1999 soon after Erlang was released as open source Experts at building bespoke scalable, high availability, high performance systems

21 Founded in 1999 soon after Erlang was released as open source Experts at building bespoke scalable, high availability, high performance systems Only company of its kind totally focused on Erlang and Erlang community

22 Founded in 1999 soon after Erlang was released as open source Experts at building bespoke scalable, high availability, high performance systems Only company of its kind totally focused on Erlang and Erlang community Over 300 clients.

23 Founded in 1999 soon after Erlang was released as open source Experts at building bespoke scalable, high availability, high performance systems Only company of its kind totally focused on Erlang and Erlang community Over 300 clients. Headquartered in London, U.K.

24 Founded in 1999 soon after Erlang was released as open source Experts at building bespoke scalable, high availability, high performance systems Only company of its kind totally focused on Erlang and Erlang community Over 300 clients. Headquartered in London, U.K. Offices: Stockholm, Krakow, Copenhagen, Aarhus, Budapest, Seattle and Zurich

25 Founded in 1999 soon after Erlang was released as open source Experts at building bespoke scalable, high availability, high performance systems Only company of its kind totally focused on Erlang and Erlang community Over 300 clients. Headquartered in London, U.K. Offices: Stockholm, Krakow, Copenhagen, Aarhus, Budapest, Seattle and Zurich Organically growing and continually investing in R & D

26 Founded in 1999 soon after Erlang was released as open source Experts at building bespoke scalable, high availability, high performance systems Only company of its kind totally focused on Erlang and Erlang community Over 300 clients. Headquartered in London, U.K. Offices: Stockholm, Krakow, Copenhagen, Aarhus, Budapest, Seattle and Zurich Organically growing and continually investing in R & D

27 Core focus

28 Speed to market Core focus

29 Core focus Speed to market Low lifetime cost

30 Core focus Speed to market Low lifetime cost Extreme reliability

31 Core focus Speed to market Low lifetime cost Extreme reliability True scalability

32 Core focus Speed to market Low lifetime cost Extreme reliability True scalability

33 University Relations

34 Some Customers

35 Some Customers 19,000,000,000 reasons to use Erlang

36 WhatsApp Real-time Messaging Text and Pictures Group chat

37 WhatsApp Numbers

38 WhatsApp Numbers 10 Erlang engineers

39 WhatsApp Numbers 10 Erlang engineers ~500M monthly users

40 WhatsApp Numbers 10 Erlang engineers ~500M monthly users 19B msg/day in / 40B msg/day out

41 WhatsApp Numbers 10 Erlang engineers ~500M monthly users 19B msg/day in / 40B msg/day out 147M concurrent connections

42 WhatsApp Numbers 10 Erlang engineers ~500M monthly users 19B msg/day in / 40B msg/day out 147M concurrent connections peak: 324K msg/s in / 712K msg/s out

43 WhatsApp Numbers 10 Erlang engineers ~500M monthly users 19B msg/day in / 40B msg/day out 147M concurrent connections peak: 324K msg/s in / 712K msg/s out

44 WhatsApp Hardware ~550 servers 2x2690v2 Ivy Bridge 10-core (40 threads total) GB RAM SSD >11,000 cores

45 WhatsApp Software FreeBSD 9.2 Erlang R16B01 (with patches)

46 Other Users of Erlang

47 Bet365

48 Bet365 Resarch team Re-did a Java system in Erlang as a POC

49 Bet365 Resarch team Re-did a Java system in Erlang as a POC Results: 5x connected users in load test 4x rate of data change Better utilisation of CPU resources

50 Bet365 Resarch team Re-did a Java system in Erlang as a POC Results: 5x connected users in load test 4x rate of data change Better utilisation of CPU resources How to convince the developers to switch?

51 Bet365 showed that Erlang

52 Bet365 showed that Erlang makes programming fun

53 Bet365 showed that Erlang makes programming fun scales and is reliable

54 Bet365 showed that Erlang makes programming fun scales and is reliable has enough depth to be interesting

55 Bet365 showed that Erlang makes programming fun scales and is reliable has enough depth to be interesting solves difficult problems with simple code

56 Bet365 showed that Erlang makes programming fun scales and is reliable has enough depth to be interesting solves difficult problems with simple code Results:

57 Bet365 showed that Erlang makes programming fun scales and is reliable has enough depth to be interesting solves difficult problems with simple code Results: production teams quickly started to appreciate the benefits of Erlang

58 Bet365 showed that Erlang makes programming fun scales and is reliable has enough depth to be interesting solves difficult problems with simple code Results: production teams quickly started to appreciate the benefits of Erlang did not want to go back to Java

59 Bet365 showed that Erlang makes programming fun scales and is reliable has enough depth to be interesting solves difficult problems with simple code Results: production teams quickly started to appreciate the benefits of Erlang did not want to go back to Java

60 Riak from Basho Distributed NoSQL database Dynamo inspired key/value store Erlang & C/C++ Powers Rovio, Danish Health Services and many more

61 Erlang History

62 There are two ways of constructing a software design:

63 There are two ways of constructing a software design: One way is to make it so simple that there are obviously no deficiencies

64 There are two ways of constructing a software design: One way is to make it so simple that there are obviously no deficiencies and the other way is to make it so complicated that there are no obvious deficiencies.

65 There are two ways of constructing a software design: One way is to make it so simple that there are obviously no deficiencies and the other way is to make it so complicated that there are no obvious deficiencies. - C.A.R. Hoare

66 Erlang s Original Requirements

67 Erlang s Original Requirements Large scale concurrency

68 Erlang s Original Requirements Large scale concurrency Soft real-time

69 Erlang s Original Requirements Large scale concurrency Soft real-time Distributed systems

70 Erlang s Original Requirements Large scale concurrency Soft real-time Distributed systems Hardware interaction

71 Erlang s Original Requirements Large scale concurrency Soft real-time Distributed systems Hardware interaction Very large software systems

72 Erlang s Original Requirements Large scale concurrency Soft real-time Distributed systems Hardware interaction Very large software systems Complex functionality

73 Erlang s Original Requirements Large scale concurrency Soft real-time Distributed systems Hardware interaction Very large software systems Complex functionality Continuous operation for many years

74 Erlang s Original Requirements Large scale concurrency Soft real-time Distributed systems Hardware interaction Very large software systems Complex functionality Continuous operation for many years Software maintenance on-the-fly

75 Erlang s Original Requirements Large scale concurrency Soft real-time Distributed systems Hardware interaction Very large software systems Complex functionality Continuous operation for many years Software maintenance on-the-fly High quality and reliability

76 Erlang s Original Requirements Large scale concurrency Soft real-time Distributed systems Hardware interaction Very large software systems Complex functionality Continuous operation for many years Software maintenance on-the-fly High quality and reliability Fault tolerance Bjarne Däcker s Licentiate Thesis:

77 wanted

78 wanted productivity

79 wanted productivity no down-time

80 wanted productivity no down-time something that always works

81 wanted

82 wanted money

83 wanted money money

84 wanted money money money

85 wanted money money money it s a rich man s world!

86 wanted money money money it s a rich man s world!

87 General vs Domain Specific Telecom

88 General vs Domain Specific Telecom C++/Java

89 General vs Domain Specific Telecom C++/Java

90 General vs Domain Specific Telecom C++/Java Erlang

91 General vs Domain Specific Telecom C++/Java Erlang

92 General vs Domain Specific Telecom C++/Java Erlang Smaller gap = money!

93 The Sweet Spot GUI Middleware Coordination Control Drivers

94 If the glove fits... needs/fit drivers coordination GUI

95 If the glove fits... needs/fit Telecom drivers coordination GUI

96 If the glove fits... needs/fit Telecom C drivers coordination GUI

97 If the glove fits... needs/fit Telecom C Erlang drivers coordination GUI

98 If our basic tool, the language in which we design and code our programs, is also complicated, the language itself becomes part of the problem rather than part of its solution.! - C.A.R. Hoare

99 Good Erlang Domains

100 Good Erlang Domains Low latency over throughput

101 Good Erlang Domains Low latency over throughput Stateful (in contrast to being stateless)

102 Good Erlang Domains Low latency over throughput Stateful (in contrast to being stateless) Massively concurrent

103 Good Erlang Domains Low latency over throughput Stateful (in contrast to being stateless) Massively concurrent Distributed

104 Good Erlang Domains Low latency over throughput Stateful (in contrast to being stateless) Massively concurrent Distributed Fault tolerant

105 Good Erlang Domains Low latency over throughput Stateful (in contrast to being stateless) Massively concurrent Distributed Fault tolerant Uses OTP

106 Good Erlang Domains Low latency over throughput Stateful (in contrast to being stateless) Massively concurrent Distributed Fault tolerant Uses OTP Non-stop operation

107 Good Erlang Domains Low latency over throughput Stateful (in contrast to being stateless) Massively concurrent Distributed Fault tolerant Uses OTP Non-stop operation Under load, Erlang programs usually performs as well as programs in other languages, often way better. Jesper Louis Andersen

108 Other Erlang Domains

109 Messaging - XMPP et al Other Erlang Domains

110 Messaging - XMPP et al Other Erlang Domains ejabberd, MongooseIM

111 Messaging - XMPP et al ejabberd, MongooseIM Webservers Other Erlang Domains

112 Other Erlang Domains Messaging - XMPP et al ejabberd, MongooseIM Webservers Yaws, Chicago Boss, Cowboy

113 Other Erlang Domains Messaging - XMPP et al ejabberd, MongooseIM Webservers Yaws, Chicago Boss, Cowboy Payment switches & soft switches

114 Other Erlang Domains Messaging - XMPP et al ejabberd, MongooseIM Webservers Yaws, Chicago Boss, Cowboy Payment switches & soft switches Vocalink, OpenFlow/LINC

115 Other Erlang Domains Messaging - XMPP et al ejabberd, MongooseIM Webservers Yaws, Chicago Boss, Cowboy Payment switches & soft switches Vocalink, OpenFlow/LINC Distributed Databases

116 Other Erlang Domains Messaging - XMPP et al ejabberd, MongooseIM Webservers Yaws, Chicago Boss, Cowboy Payment switches & soft switches Vocalink, OpenFlow/LINC Distributed Databases Riak, CouchDB, Scalaris

117 Other Erlang Domains Messaging - XMPP et al ejabberd, MongooseIM Webservers Yaws, Chicago Boss, Cowboy Payment switches & soft switches Vocalink, OpenFlow/LINC Distributed Databases Riak, CouchDB, Scalaris Queueing systems

118 Other Erlang Domains Messaging - XMPP et al ejabberd, MongooseIM Webservers Yaws, Chicago Boss, Cowboy Payment switches & soft switches Vocalink, OpenFlow/LINC Distributed Databases Riak, CouchDB, Scalaris Queueing systems RabbitMQ (AMQP)

119 The glove fits! Low latency Stateful Massively concurrent Distributed Fault tolerant Messaging Webservers Soft switches Distributed DBs Queueing systems

120 The Golden Trinity Of Erlang

121 To Share Or Not To Share

122 To Share Or Not To Share Memory

123 To Share Or Not To Share Memory P1

124 To Share Or Not To Share Memory P1 P2

125 To Share Or Not To Share Memory Corrupt P2

126 To Share Or Not To Share Memory Corrupt

127 To Share Or Not To Share Memory Corrupt Memory

128 To Share Or Not To Share Memory Corrupt Memory P1

129 To Share Or Not To Share Memory Corrupt Memory Memory P1

130 To Share Or Not To Share Memory Corrupt Memory Memory P1 P2

131 To Share Or Not To Share Memory Corrupt Corrupt Memory P2

132 To Share Or Not To Share Memory Corrupt Memory P2

133 Message Passing P1 M P2 P1 sends M to P2.

134 Message Passing P1 M P2 P1 sends M to P2. Every process has a mailbox

135 Message Passing P1 M P2 P1 sends M to P2. Every process has a mailbox Messages are received: receive {tag, Value} -> Value; N when is_integer(n) -> N + 42 end

136 Failures Anything that can go wrong, will go wrong Murphy

137 Failures Anything that can go wrong, Programming errors will go wrong Murphy

138 Failures Anything that can go wrong, Programming errors will go wrong Disk failures Murphy

139 Failures Anything that can go wrong, Programming errors will go wrong Disk failures Murphy Network failures

140 Failures Anything that can go wrong, Programming errors will go wrong Disk failures Murphy Network failures Most programming paradigmes are fault in-tolerant

141 Failures Anything that can go wrong, Programming errors will go wrong Disk failures Murphy Network failures Most programming paradigmes are fault in-tolerant must deal with all errors or die

142 Failures Anything that can go wrong, Programming errors will go wrong Disk failures Murphy Network failures Most programming paradigmes are fault in-tolerant must deal with all errors or die

143 Failures Anything that can go wrong, Programming errors will go wrong Disk failures Murphy Network failures Most programming paradigmes are fault in-tolerant must deal with all errors or die Erlang is fault tolerant by design

144 Failures Anything that can go wrong, Programming errors will go wrong Disk failures Murphy Network failures Most programming paradigmes are fault in-tolerant must deal with all errors or die Erlang is fault tolerant by design failures are embraced and managed

145 Failures Anything that can go wrong, Programming errors will go wrong Disk failures Murphy Network failures Most programming paradigmes are fault in-tolerant must deal with all errors or die Erlang is fault tolerant by design failures are embraced and managed source: should-we-be-worried-about-this-brooklyn-measles-outbreak source:

146 Let It Fail convert(monday) -> 1;! convert(tuesday) -> 2;! convert(wednesday) -> 3;! convert(thursday) -> 4;! convert(friday) -> 5;! convert(saturday) -> 6;! convert(sunday) -> 7!;! convert(_) ->! {error, unknown_day}.!

147 Let It Fail convert(monday) -> 1;! convert(tuesday) -> 2;! convert(wednesday) -> 3;! convert(thursday) -> 4;! convert(friday) -> 5;! convert(saturday) -> 6;! convert(sunday) -> 7!.!!

148 Let It Fail convert(monday) -> 1;! convert(tuesday) -> 2;! convert(wednesday) -> 3;! convert(thursday) -> 4;! convert(friday) -> 5;! convert(saturday) -> 6;! convert(sunday) -> 7!.!! Erlang encourages offensive programming

149 Handling Failure P1 P2 P1 monitors P2. P1 P2 P1 and P2 are linked.

150 Intentional Programming a style of programming where the reader of a program can easily see what the programmer intended by their code. [1] [1]

151 Intentional Dictionary data retrieval - dict:fetch(key, Dict) = Val EXIT the programmer knows a specific key should be in the dictionary and it is an error if it is not. search - dict:find(key, Dict) = {ok, Val} error. it is unknown if the key is there or not and both cases must be dealt with. test - dict:is_key(key, Dict) = Boolean knowing if a key is present is enough.

152 Benefits of let-it-fail Data Mobility component breakdown Source: h*p:// JanHenryNystrom/produc;vity- gains- in- erlang

153 Benefits of let-it-fail Data Mobility component breakdown Source: h*p:// JanHenryNystrom/produc;vity- gains- in- erlang code that solves the problem

154 Benefits of let-it-fail Data Mobility component breakdown Source: h*p:// JanHenryNystrom/produc;vity- gains- in- erlang code that solves the problem

155 Benefits of let-it-fail Data Mobility component breakdown Source: h*p:// JanHenryNystrom/produc;vity- gains- in- erlang code that solves the problem

156 Benefits of let-it-fail Data Mobility component breakdown Source: h*p:// JanHenryNystrom/produc;vity- gains- in- erlang code that solves the problem

157 Benefits of let-it-fail Data Mobility component breakdown Source: h*p:// JanHenryNystrom/produc;vity- gains- in- erlang code that solves the problem

158 Benefits of let-it-fail Data Mobility component breakdown Source: h*p:// JanHenryNystrom/produc;vity- gains- in- erlang code that solves the problem

159 Benefits of let-it-fail Data Mobility component breakdown Source: h*p:// JanHenryNystrom/produc;vity- gains- in- erlang code that solves the problem 3x

160 Productivity with Motorola Goal: Development of a mission-critical telecom gateway for TETRA To be developed from scratch using Erlang and some drivers in C The gateway translates between proprietary protocols and the ISI standard Developed using a two six man team over four year A total of 72 staff months used on the work

161 Productivity with Motorola Function point analysis Language agnostic measurement of problem size input input input output output output input Internal storage

162 Show me the money! Function Point Analysis of the size of the problem Conservative estimation of the number of inputs, outputs and internal storage Includes design, box test, system test, project management efforts

163 Isolate Errors

164 PROPAGATING EXIT SIGNALS PidA PidB PidC

165 PROPAGATING EXIT SIGNALS PidA PidB PidC

166 PROPAGATING EXIT SIGNALS {'EXIT', PidA, Reason} PidB PidC

167 PROPAGATING EXIT SIGNALS PidB PidC

168 PROPAGATING EXIT SIGNALS {'EXIT', PidB, Reason} PidC

169 PROPAGATING EXIT SIGNALS

170 TRAPPING AN EXIT SIGNAL PidA PidB PidC

171 TRAPPING AN EXIT SIGNAL PidA PidB PidC

172 TRAPPING AN EXIT SIGNAL {'EXIT', PidA, Reason} PidB PidC

173 TRAPPING AN EXIT SIGNAL PidB PidC

174 Supervision Trees supervisor supervisor worker worker worker worker worker The OTP library is built on this principle Erlang Solutions Ltd. 45

175 The OTP Supervisor Specifies a default restart strategy supervisor one_for_one one_for_all rest_for_one simple_one_for_one worker worker worker Child spec for how a child is restarted permanent transient temporary Erlang Solutions Ltd. 46

176 Supervisors

177 Simple Manager/Worker Pattern

178 Realities of software development???? Source:

179 Realities of software development Product Owner Source:

180 Business benefits of supervisors

181 Business benefits of supervisors Only one process dies

182 Business benefits of supervisors Only one process dies isolation gives continuous service

183 Business benefits of supervisors Only one process dies isolation gives continuous service Everything is logged

184 Business benefits of supervisors Only one process dies isolation gives continuous service Everything is logged you know what is wrong

185 Business benefits of supervisors Only one process dies isolation gives continuous service Everything is logged you know what is wrong Corner cases can be fixed at leisure

186 Business benefits of supervisors Only one process dies isolation gives continuous service Everything is logged you know what is wrong Corner cases can be fixed at leisure Product owner in charge!

187 Business benefits of supervisors Only one process dies isolation gives continuous service Everything is logged you know what is wrong Corner cases can be fixed at leisure Product owner in charge! Not the software!

188 Business benefits of supervisors Only one process dies isolation gives continuous service Everything is logged you know what is wrong Corner cases can be fixed at leisure Product owner in charge! Not the software! Software architecture that supports iterative development

189 Preparing for battle Source:

190 When do I get my ROI? Speed Proto-typing Development Live

191 When do I get my ROI? Speed Some Cool Technology Proto-typing Development Live

192 When do I get my ROI? Speed Erlang Some Cool Technology Proto-typing Development Live

193 When do I get my ROI? Speed Proto-typing Development Live

194 When do I get my ROI? Speed Erlang Proto-typing Development Live

195 When do I get my ROI? Speed Erlang Some Cool Technology Proto-typing Development Live

196 Erlang/OTP

197 Middleware

198 Design Patterns Middleware

199 Middleware Design Patterns Fault Tolerance

200 Middleware Design Patterns Fault Tolerance Distribution

201 Middleware Design Patterns Fault Tolerance Distribution Upgrades

202 Middleware Design Patterns Fault Tolerance Distribution Upgrades Packaging

203 Middleware Design Patterns Fault Tolerance Distribution Upgrades Packaging

204 OTP Tools

205 Development OTP Tools

206 OTP Tools Development Test Frameworks

207 OTP Tools Development Test Frameworks Release & Deployment

208 OTP Tools Development Test Frameworks Release & Deployment Debugging & Monitoring

209 OTP Benefits

210 Less Code OTP Benefits

211 OTP Benefits Less Code Less Bugs

212 OTP Benefits Less Code Less Bugs More Solid Code

213 OTP Benefits Less Code Less Bugs More Solid Code More Tested Code More Free Time

214 Behaviours GENERIC BEHAVIOUR MODULE Server process SPECIFIC CALLBACK MODULE

215 OTP Behaviours

216 OTP Behaviours Servers

217 OTP Behaviours Servers Finite State Machines

218 OTP Behaviours Servers Finite State Machines Event Handlers

219 OTP Behaviours Servers Finite State Machines Event Handlers Supervisors

220 OTP Behaviours Servers Finite State Machines Event Handlers Supervisors Applications

221 Nasty Things Handled by OTP

222 Nasty Things Handled by OTP Who are you gonna call?

223 Client Server

224 {request, Pid, Message} Client Server call(name, Message) ->! Name! {request, self(), Message},! receive! {reply, Reply} -> Reply! end.

225 {request, Pid, Message} Client {reply, Reply} Server call(name, Message) ->! Name! {request, self(), Message},! receive! {reply, Reply} -> Reply! end. reply(pid, Reply) ->! Pid! {reply, Reply}.

226 {request, Pid, Message} Client {reply, Reply} Server

227 {request, Pid, Message} Client {reply, Reply} Server Server 2

228 {request, Pid, Message} Client {reply, Reply} Server {reply, Reply} Server 2

229 {request, Pid, Message} Client {reply, Reply} Server {reply, Reply} Server 2 call(name, Msg) ->! Ref = make_ref(),! Name! {request, {Ref, self()}, Msg},! receive {reply, Ref, Reply} -> Reply end.!! reply({ref, Pid}, Reply) ->! Pid! {reply, Ref, Reply}.

230 {request, {Ref, self()}, Message} Client {reply, Ref, Reply} Server {reply, Reply} Server 2 call(name, Msg) ->! Ref = make_ref(),! Name! {request, {Ref, self()}, Msg},! receive {reply, Ref, Reply} -> Reply end.!! reply({ref, Pid}, Reply) ->! Pid! {reply, Ref, Reply}.

231 {request, {Ref, self()}, Message} Client {reply, Ref, Reply} Server {reply,???, Reply} Server 2 call(name, Msg) ->! Ref = make_ref(),! Name! {request, {Ref, self()}, Msg},! receive {reply, Ref, Reply} -> Reply end.!! reply({ref, Pid}, Reply) ->! Pid! {reply, Ref, Reply}.

232 PidA PidB

233 {request, {Ref, PidA}, Msg} PidA PidB

234 PidA

235 PidA call(name, Msg) ->! Ref = erlang:monitor(process, Name),! Name! {request, {Ref, self()}, Msg},! receive!! {reply, Ref, Reply} ->!! erlang:demonitor(ref),!! Reply;!! {'DOWN', Ref, process, _Name, _Reason} ->!! {error, no_proc}! end.

236 PidA PidB

237 {request, {Ref, PidA}, Msg} PidA PidB

238 PidA {reply, Ref, Reply} PidB

239 PidA {'DOWN', Ref, process, PidB, Reason}

240 {'DOWN', Ref, process, PidB, Reason} PidA call(name, Msg) ->! Ref = erlang:monitor(process, Name),! Name! {request, {Ref, self()}, Msg},! receive!! {reply, Ref, Reply} ->!! erlang:demonitor(ref, [flush]),!! Reply;!! {'DOWN', Ref, process, _Name, _Reason} ->!! {error, no_proc}! end.

241 Dealing with deadlocks

242 Dealing with deadlocks 7 years of coding Erlang

243 Dealing with deadlocks 7 years of coding Erlang Time spent on deadlock issues.

244 Dealing with deadlocks 7 years of coding Erlang Time spent on deadlock issues. 1 hour (due to lack of experience with OTP)

245 Selling Others on Erlang

246 In Theory... You conquer the TALC group by group in one smooth motion Erlang Solutions Ltd. 66

247 In Theory... You conquer the TALC group by group in one smooth motion Erlang Solutions Ltd. 66

248 In Theory... You conquer the TALC group by group in one smooth motion Erlang Solutions Ltd. 66

249 In Theory... You conquer the TALC group by group in one smooth motion Erlang Solutions Ltd. 66

250 In Theory... You conquer the TALC group by group in one smooth motion Erlang Solutions Ltd. 66

251 In Theory... You conquer the TALC group by group in one smooth motion Erlang Solutions Ltd. 66

252 In Theory... You conquer the TALC group by group in one smooth motion but it s just an illusion :-( Erlang Solutions Ltd. 66

253 Cracks and a Chasm Erlang Solutions Ltd. 67

254 Cracks and a Chasm Technology enthusiasts Erlang Solutions Ltd. 67

255 Cracks and a Chasm Visionaries Technology enthusiasts Erlang Solutions Ltd. 67

256 Cracks and a Chasm Pragmatists Visionaries Technology enthusiasts Erlang Solutions Ltd. 67

257 Cracks and a Chasm Pragmatists Conservatives Visionaries Technology enthusiasts Erlang Solutions Ltd. 67

258 Cracks and a Chasm Pragmatists Conservatives Technology enthusiasts Visionaries Laggards Erlang Solutions Ltd. 67

259 Cracks and a Chasm Crack Pragmatists Conservatives Technology enthusiasts Visionaries Laggards Erlang Solutions Ltd. 67

260 Cracks and a Chasm Crack Crack Pragmatists Conservatives Technology enthusiasts Visionaries Laggards Erlang Solutions Ltd. 67

261 Cracks and a Chasm Chasm Crack Crack Pragmatists Conservatives Technology enthusiasts Visionaries Laggards Erlang Solutions Ltd. 67

262 The Chasm

263 The Chasm From Visionaries to Pragmatists

264 The Chasm From Visionaries to Pragmatists Visionaries buy a change agent to get a radical discontinuity

265 The Chasm From Visionaries to Pragmatists Visionaries buy a change agent to get a radical discontinuity Pragmatists want a productivity improvement for existing operations

266 The Chasm From Visionaries to Pragmatists Visionaries buy a change agent to get a radical discontinuity Pragmatists want a productivity improvement for existing operations Pragmatists want evolution, not revolution

267 The Chasm From Visionaries to Pragmatists Visionaries buy a change agent to get a radical discontinuity Pragmatists want a productivity improvement for existing operations Pragmatists want evolution, not revolution Pragmatists wants references

268 Visionaries vs Pragmatists Visionaries have four characteristics that alienate Pragmatists: Erlang Solutions Ltd. 69

269 Visionaries vs Pragmatists Visionaries have four characteristics that alienate Pragmatists: Lack of respect for their colleagues experiences Erlang Solutions Ltd. 69

270 Visionaries vs Pragmatists Visionaries have four characteristics that alienate Pragmatists: Lack of respect for their colleagues experiences Takes greater interest in technology than in their industry Erlang Solutions Ltd. 69

271 Visionaries vs Pragmatists Visionaries have four characteristics that alienate Pragmatists: Lack of respect for their colleagues experiences Takes greater interest in technology than in their industry Fail to recognise the importance of existing product infrastructure Erlang Solutions Ltd. 69

272 Visionaries vs Pragmatists Visionaries have four characteristics that alienate Pragmatists: Lack of respect for their colleagues experiences Takes greater interest in technology than in their industry Fail to recognise the importance of existing product infrastructure Overall disruptiveness Erlang Solutions Ltd. 69

273 Whole Product Planning Simplified for chasm crossing Training & Support Standards & Procedures Generic Product Additional Software Additional Hardware Anything else to achieve your compelling reason to buy Installation & Debugging System Integration Erlang Solutions Ltd. 70

274 Whole Product Planning Simplified for chasm crossing Training & Support Standards & Procedures Generic Product Additional Software Additional Hardware Anything else to achieve your compelling reason to buy = marketing promise to win the sale Installation & Debugging System Integration Erlang Solutions Ltd. 70

275 Erlang Whole Product 1/2 Generic product: Erlang compiler and runtime Additional software: rich library shipped with each release many open source libraries Training & support: ESL provides many courses ErlangCamp also provides training Erlang Solutions Ltd. 71

276 Erlang Whole Product 2/2 System integration: mostly case-by-case few public success stories Installation & Debugging: Adequate functionalities for installing applications Wombat: ESL tool for Operations and Maintenance Good debugging tools, but not well publicised Erlang Solutions Ltd. 72

277 The EuroVision Link Or why Java is used more than Erlang source: source: source: v=1&c=iwsasset&k=2&d=x7wjla88cweo9hktrlanxrtyxdj3bsapheprrdcrbmiaeiszj73mxej82t5ovccz

278 The EuroVision Link Or why Java is used more than Erlang Lots of wrapping source: source: source: v=1&c=iwsasset&k=2&d=x7wjla88cweo9hktrlanxrtyxdj3bsapheprrdcrbmiaeiszj73mxej82t5ovccz

279 The EuroVision Link Or why Java is used more than Erlang Lots of wrapping The real McCoy source: source: source: v=1&c=iwsasset&k=2&d=x7wjla88cweo9hktrlanxrtyxdj3bsapheprrdcrbmiaeiszj73mxej82t5ovccz

280 Source:

281 Learning Erlang

282 Learning Erlang ESL training courses

283 Learning Erlang ESL training courses Learn You Some Erlang

284 Learning Erlang ESL training courses Learn You Some Erlang Use the erlang-questions mailing list

285 Learning Erlang ESL training courses Learn You Some Erlang Use the erlang-questions mailing list Do it hands-on

286 Learning Erlang ESL training courses Learn You Some Erlang Use the erlang-questions mailing list Do it hands-on Give it time to sink in!!!

287 Elixir

288 Elixir Built on top of the Erlang VM

289 Elixir Built on top of the Erlang VM More Ruby-like syntax

290 Elixir Built on top of the Erlang VM More Ruby-like syntax Hygienic macros - easy to do DSLs

291 Elixir Built on top of the Erlang VM More Ruby-like syntax Hygienic macros - easy to do DSLs But you still have to learn the Erlang programming model

292 Key building blocks

293 Key building blocks Share nothing processes

294 Key building blocks Share nothing processes Message passing

295 Key building blocks Share nothing processes Message passing Fail fast approach

296 Key building blocks Share nothing processes Message passing Fail fast approach Link/monitor concept

297 Key building blocks Share nothing processes Message passing Fail fast approach Link/monitor concept You can deal with failures in a sensible manner because you have a language for them.

298 Cruising with Erlang

299 Cruising with Erlang Understand the failure model

300 Cruising with Erlang Understand the failure model Embrace failure!

301 Cruising with Erlang Understand the failure model Embrace failure! Use supervision patterns to deliver business value

302 Cruising with Erlang Understand the failure model Embrace failure! Use supervision patterns to deliver business value Stay in charge!

303 Cruising with Erlang Understand the failure model Embrace failure! Use supervision patterns to deliver business value Stay in charge!