пятница, 31 августа 2018 г.

5 за 5 (история 9)

Вот не поймешь, хвалят коллеги или ругаются... (с)
1. "Punishing failure makes it worse. Unshared failures are the experience of one individual. They are not institutional learning." ( by Jessica Kerr tweet)

2. The problem with root cause analysis is there’s always going to be another root cause.
Robustness happens at a higher level. And resilience at a higher level still. (by Jessica Kerr tweet)

3. Очередная история про Git от Сергея Сергеева, беседа в подкасте "Подлодка". Напомню ссылку на отличную лекцию Сергея.

4. "Тимлид — это сержант в IT-подразделении" - хорошее интервью с Романом Ивлиевым.

5. Неплохая панелька тимлидов с инсайдами из известных компаний.  Отлично смотрится на скорости х1.5


пятница, 17 августа 2018 г.

5 за 5 (история 8)

И снова с вами рубрика "что интересного было в ленте на этой неделе".

1. Building and Testing Resilient Web Applications with Toxiproxy.
Статья, видео.
"A resilient system is one that functions with one or more components being unavailable or unacceptably slow. Applications quickly become intertwined with their external services if not carefully monitored, leading to minor dependencies becoming single points of failure.
For example, the only part of Shopify that relies on the session store is user sign-in - if the session store is unavailable, customers can still purchase products as guests. Any other behaviour would be an unfortunate coupling of components. This post is an overview of the tools and techniques we used to make Shopify more resilient in preparation for the holiday season."

2. What is Soak Testing?
"Soak testing (otherwise known as endurance testing, capacity testing, or longevity testing) involves testing the system to detect the performance-related issues such as stability and response time by requesting the designed load on a system."

3. "Элита" - хорошая статья про программистов и их заблуждения.
"...программисты — обычные головой люди, их элитарность самодутая и не стоит ни гроша. Не самое приятное, что входило мне в мозг, признаюсь. Но и закрывать глаза уже не получается."
И сам блог, кстати, рекомендую. Интересные статьи.

4. Тред в твиттере про то, что для влияния инженеру не нужно становиться менеджером.

5. Небольшой тред-слайды в твиттере "Chaos Engineering is about engineering around the chaos inherent in the system".



пятница, 10 августа 2018 г.

5 за 5 (истории 6-7)

Последняя неделя отпуска прошла в режиме "без связи", поэтому предыдущий выпуск пропустил.
В этом нагоняем, продолжая читать "Just Enough Software Architecture" и добавив интересных ссылок.

1. Работа команды над решением задачи снижения риска ухудшения архитектуры путем прояснения текущего ее состояния для новых членов команды:
we were aware of providing coverage of the three primary modelsthe domain, the design, and the code models —and also the three primary architectural viewtypesthe module, runtime, and allocation views.
We started with the easiest documentation to produce and gradually added in more expensive parts. After each one, we asked ourselves if the risk had substantially reduced and we calibrated that evaluation based on our coverage of the viewtypes and models. When possible, we built representative and textual models rather than fully general and graphical ones. We decided to create a graphical model of our modules and component assembly since they were relatively easy to produce and conveyed more information than our textual models. We stopped when we had covered the primary models and viewtypes, trusting that the new developers would be able to use what we provided as a skeleton of understanding and would hang detailed knowledge from it.

2. Risk-driven approach to software architecture consists of identifying risks, deciding the best set of techniques to mitigate the risks, and then evaluating the remaining risk...
Instead of applying architecture techniques until we ran out of them, until the documentation binder was complete, or until the project was canceled, we regularly re-evaluated the remaining risks and stopped when they had subsided.

3. The core of architectural understanding is to be able to get at the “why” questions. Having the understanding does not mean you must follow a certain process, or program in a certain language, or write diagrams on paper. Understanding software architecture means that you have internalised the (admittedly incomplete and imperfect) knowledge and abstractions that have been built up, and that you can apply that understanding when building new systems or analysing existing ones.

Дальше книга не идет :( Отложил пока в сторону.



Новости:
1.  Анонс Heisenbug 2018 Moscow. Ждем докладчиков, и можно уже планировать билеты и поездку.

2. Отличная статья про машинное обучение для тех, кто не знает, но стеснялся спросить: "Машинное обучение для людей. Разбираемся простыми словами".

3. Андрей Сатарин зафигачил хороший тред про тестирование.

4. How to measure exploratory tests?

5. Мысли вслух: институт team lead-ов в компании очень важная составляющая ее работы. Их нужно растить, искать, развивать. Без этого все очень сложно получается. Но, млин, так редко на моей памяти это происходит. Часто, а развитие чаще всего, это ложится на плечи самого лида. В целом то оно и правильно, но с помощью всегда лучше получается обычно. Если кому интересно, то тут вот в Питере правильная конфа будет в сентябре "Профессиональная конференция про тимлидов и для тимлидов".






пятница, 27 июля 2018 г.

5 за 5 (история 5)

Отпускное чтиво, навеянное чередой не связанных между собой событий, но приведших к одинаковым мыслям.
Заметки из книги "Just Enough Software Architecture".

1. Software architecture is about the design of your system and the impact it has on the system’s qualities, qualities like performance, security, and modifiability.  This definition discusses how architecture differs from detailed design and how some of your biggest design decisions can have implications deep into the code.

2. 3 типа архитектурного подхода:
Imagine that your performance requirements say that your system must respond to requests within 50ms.
Here are some possible ways that you could approach the system’s architecture, given the three design approaches:
• Architecture-indifferent design. If you followed architecture-indifferent design, you could copy the distributed processing architecture from your last system and discover, hopefully not too late, that its inter-machine messaging over head eats up most of that 50ms, leaving little time to do the real processing. To succeed, you either change the architecture or write very efficient code that can complete in 10ms.
• Architecture-focused design. If you followed architecture-focused design, you would deliberately choose an architecture that is compatible with that requirement, such as a client-server architecture. The single remote call to the server might take 10ms, which leaves you a reasonable 40ms to do the real processing.
• Architecture hoisting. If you hoist the performance goal into the architecture, you would ask yourself how the architecture could ensure that a 50ms response was always achievable. Perhaps your investigation reveals that there are peak demand times that could overload your servers, so you build software to recruit additional processing, perhaps from a cloud of servers.

3. How much design and architecture should developers do?
There is much active debate about this question and several kinds of answers have been suggested:
No up-front design. Developers should just write code. Design happens, but is coincident with coding, and happens at the keyboard rather than in advance.
Use a yardstick. For example, developers should spend 10% of their time on architecture and design, 40% on coding, 20% on integrating, and 30% on testing.
Build a documentation package. Developers should employ a comprehensive set of design and documentation techniques sufficient to produce a complete
written design document.
Ad hoc. Developers should react to the project needs and decide on the spot how much design to do.

4. The risk-driven model guides developers to apply a minimal set of architecture techniques to reduce their most pressing risks. It suggests a relentless questioning process: “What are my risks? What are the best techniques to reduce them? Is the risk mitigated and can I start (or resume) coding?”
The risk-driven model can be summarized in three steps:
1. Identify and prioritize risks
2. Select and apply a set of techniques
3. Evaluate risk reduction

5. Прочитал первые 3 главы. Пока не впечатлен. Продолжение следует. Дальше идут примеры. Посмотрим, как зайдет. 

пятница, 20 июля 2018 г.

5 за 5 (история 4)

Привет. С вами снова очередные полезности за неделю. Признаться, свежих было немного, поэтому, пошарившись по архивам, закину вам чуток из прошлого.

1. Harvest, Yield, and Scalable Tolerant Systems (PDF)
Обычно мне тяжело даются такие "около академические" труды, CAP-теорема и вот это все. Но тут хорошо зашло: новые термины для того, что уже раньше использовалось в работе и обозначалось "на пальцах".

We assume that clients make queries to servers, in which case there are at least two metrics for correct behaviour: yield, which is the probability of completing a request, and harvest, which measures the fraction of the data reflected in the response, i.e. the completeness of the answer to the
query. Yield is the common metric and is typically measured in “nines”: “four-nines availability” means a completion probability of 0.9999 . In practice, good HA systems aim for four or five nines. In the presence of faults there is typically a tradeoff between providing no answer (reducing yield) and providing an imperfect answer (maintaining yield, but reducing harvest). Some applications do not tolerate harvest degradation because any deviation from the single well-defined correct behaviour renders the result useless.

2. За все время работы в разработке я часто сталкивался с верой в магическую должность "Архитектор" (Архитектор ПО, Системный архитектор и прочее). Видимо мне не повезло встретить и поработать с настоящими архитекторами, если они существуют. А сам я, IMHO, как-то хреново "архитектирую".
Тем не менее, вот вам несколько полезных ссылок про архитектуру ПО "на подумать-почитать":
3. "Программный комитет HolyJS изнутри" - подробный рассказ про процесс подготовку докладов (= содержательной части конфы) к конференции HolyJS. В нашем ПК Heisenbug конфы процессы очень похожие.

4. Немного истории из своего блога "Популярная психология в IT и не только".

5. Интересная цитата, надо книжку почитать
An Approach to Cybernetics (1961) by Gordon Pask
"Observers are men, animals, or machines able to learn about their environment and impelled to reduce their uncertainty about the events which occur in it, by dint of learning... [We] shall examine human observers who, because we have an inside understanding of their observational process, belong to a special category. For the moment, we shall not bother with HOW an observer learns, but will concentrate upon WHAT he learns about, i.e. what becomes more certain."