1

Topic: Re: Microservices, EC2, S3

Hello, Tyomchik, you wrote: Those> I Understand that , missed revolution etc. I Suggest to consider a paradigm of microservices (as it was described by Martin Fowler) in practical application. Who has an experience of purchase IaaS, adjustments of it in cluster EMR, what task scheduler was used ( or Apache Spark or something else)? I Am sorry, but such feeling that a subject created with what help that bullshit bingo or the abstract... How all named words and terms are connected in your understanding?

2

Re: Re: Microservices, EC2, S3

Hello, Tyomchik, you wrote: Those> I Understand that , missed revolution etc. I Suggest to consider a paradigm of microservices (as it was described by Martin Fowler) in practical application. Well if ... Fowler... To it was linux way app development, COM/CORBA, dbus, etc. not especially I understand that the new was offered by Fowler except reconsideration of known concepts within the limits of a bit different task. On me so microservices on the basis of Erlang+rabbitmq it is beautiful enough .

3

Re: Re: Microservices, EC2, S3

Hello, Tyomchik, you wrote: Those> in clouds, a heap of microservices, a lambda, ,  or analogs,  , I ask you, continue , do not stop...

4

Re: Re: Microservices, EC2, S3

Hello, DrDred, you wrote: DD> As all named words and terms are connected in your understanding? Are not connected in any way. Get acquainted, it is Tyomchik.

5

Re: Re: Microservices, EC2, S3

Hello, Tyomchik, you wrote: Those> I Understand that , missed revolution etc. I Suggest to consider a paradigm of microservices (as it was described by Martin Fowler) in practical application. Martin Fowler has good books, like "Architecture of corporate program applications" in which are not present words about microservices. Clear business in "UML Bases the Short manual on a unified modeling language" and other books hardly that too will be. But we have Google, and he says that it is a question of this article. My judgement that in comparison with the book "Architecture of corporate program applications" at them the short review which the one hardly understands who not in a subject there is presented, and I see it so. In our world there are objects (entities) and actions over them. The object can influence both itself, and on other object, anyway so it to themselves is represented by people. On the one hand we have nouns (essence), with another verbs (actions). Accordingly for communication it is used two approaches, it is date transmission and functions. In one case we transfer the data of the object, these are fields and other resources, in other functions, for copies of a class it there will be methods and so on. Now about what technologies were included into the first and in the second group: 1) the data: CRUD, SQL, REST 2) functions: RPC, CORBA, DCOM, SOAP And so on. To take the same HTTP, at once it is visible that it is included into the first group, but there is an important difference, we consider not generally everything that is and is anchored to principle CRUD though for the general understanding I and pushed here it, and as SQL, and that though is somehow connected to architecture of applications. Time we started talking about Martin Fowler who wrote as and about UML, is mandatory it is necessary to mention OMG. BPMN - the graphic notation for modeling business of processes. MDA - the concept it is modeling the oriented approach to software development Meta-Object Facility UML - language of the graphic description for the objective modeling XMI - the standard for an exchange of meta data by means of language XML. CWM CORBA IDL - language of the description of interfaces IIOP DDS Object Management Architecture (OMA) it would be possible to someone interesting to look at the same DDS. About DCOM and CORBA it is possible to esteem on this site in section of articles. I am far from Java and nevertheless there is such piece as servlets. We bring preliminary totals, to distinguish technologies simply enough. The first use 4 functions on CRUD, the second transfer actions which need to be carried out. And both that and another it is possible to use for achievement of the same result, that is function invocation to transfer the data or to cause functions I use resources on CRUD. But presently the balance drifts aside CRUD, and all more and more because this method guzzles less traffics and is better scaled. Plus to it still fasten decentralization. Thus any special revolution also was not, we now consider for a long time already "moss-grown" technologies. As to network interaction here enough to read articles about clusters. They as have a huge set of problems. From Martin Fowler's book "Architecture of corporate program applications" I acquired that it is not necessary to perceive all especially from the point of view of component interaction on a network. And in other book about TCP protocol it was said that the network network ,  is one speed and reliability,  another, and the Internet and at all demands absolutely other approach. It, by the way, once again reminds about clusters, a cloudlet and everyones , and in a reality efficiency of application will be influenced by a location of servers on which it is scattered. We look here: Classification of the multiprocessor computing systems And now here: Schematic type of SMP-architecture Personally I consider that it is impossible to dismiss symmetric . Let's far not walk at all, and we take DDR3. Here the standard at Intel of generation before last. At standard title DDR3-1600 Frequency of storage, MHz 12800 Time of a cycle, nanosecond 5,00 Frequencies of the bus, MHz 800 Effective (doubled) speeds, million transmissions/with 1600 Title of PC3-12800 unit Peak data flow rate at the 64-bit bus of the data in an one-channel mode, MB/WITH 12800 Look at volumes and transit velocity, monstrous productivity. And now let's recall TCP on which to us suggest to drive a data interchange and functions. No, it is possible certainly  that in the normal program all not so, there call stacks, still to recall any assembler. But if it is a question of Martin Fowler, and at it there all is sated by various concepts, very much he likes to divide applications into levels and layers where levels it and is interactions between the server and the client, and layers sharing in the application. But if we fell to such abstractions, and they there completely not sickly who told that the server and the client is mandatory should be in a network. And I drive to that if it at level SMP, for example, connection of plug-ins, let works the same as well as at level of clusters simply using foliation, and without a difference there on Fowler, on Brown or on a horse-radish from mountain. We will hope that I expressed not clearly enough and foggy that my work appreciated.

6

Re: Re: Microservices, EC2, S3

Hello, velkin, you wrote: V> we Will hope that I expressed not clearly enough and foggy that my work appreciated. Yes, " a chasm" (Unthinking Depths) to You candidate to write for some Duma members.

7

Re: Re: Microservices, EC2, S3

Hello, bnk, you wrote: bnk> to You candidate to write for some Duma members. It agree, he wrote such abstract to it, and Tyomchik put a smile in it. 100 % it is assured that at all did not read.

8

Re: Re: Microservices, EC2, S3

Hello, Kernan, you wrote: K> On me so microservices on the basis of Erlang+rabbitmq it is beautiful enough . Erlang+rabbitmq are buses which is orthogonal to microservices.

9

Re: Re: Microservices, EC2, S3

Hello, Kernan, you wrote: K> Hello, Tyomchik, you wrote: Those>> I Understand that , missed revolution etc. I Suggest to consider a paradigm of microservices (as it was described by Martin Fowler) in practical application. K> well if ... K> Fowler... To it was linux way app development, COM/CORBA, dbus, etc. not especially I understand that the new was offered by Fowler except reconsideration of known concepts within the limits of a bit different task. K> on me so microservices on the basis of Erlang+rabbitmq it is beautiful enough . And still from all remaining you missed the main difference of microservices...

10

Re: Re: Microservices, EC2, S3

Hello, Tom, you wrote: Tom> And still from all remaining you missed the main difference of microservices... No, did not miss. You are mistaken.