内容简介:By standardise common error handling, observability logic or other patterns as reusable modules, it ensures consistency of observability standards across micro-services, codebases or teams.Hooks can automate standardised log, metrics and tracing in a struc
hooks
Purpose
Consistency as Toolings (CasT)
what to observability and control patterns, like linters to code styles
By standardise common error handling, observability logic or other patterns as reusable modules, it ensures consistency of observability standards across micro-services, codebases or teams.
/* handler.js */ import logger, metrics from '@opbi/toolchain'; import { eventLogger, eventTimer } from '@opbi/hooks'; const withObserver = chain(eventLogger, eventTimer); const getSubscription = withObserver(userProfileApi.getSubscription); const cancelSubscription = withObserver(subscriptionApi.cancel) const handleUserCancelSubscription = async ({ userId }, meta, context) => { const { subscriptionId } = await getSubscription( { userId }, meta, context ); await cancelSubscription({ subscriptionId }, meta, context); }; export default withObserver(handleUserCancelSubscription); /* router.js */ import handleUserCancelSubscription from './handler.js'; await handleUserCancelSubscription({ userId }, meta, { logger, metrics });
Hooks can automate standardised log, metrics and tracing in a structure reflecting the call stacks. This greatly improves observability coverage and makes monitor and debugging a breeze with good precision locating problematic function.
[info] event: handleUserCancelSubscription [info] event: handleUserCancelSubscription.getSubscription [error] event: handleUserCancelSubscription.cancelSubscription, type: TimeoutError
Readability, Reusability, Testability (RRT)
Turn scattered repeatitive control mechanism or observability code from interwined blocks to more readable, reusable, testable ones.
By abstract out common control mechanism and observability code into well-tested, composable hooks, it can effectively half the verboseness of your code. This helps to achieve codebase that is self-explanatory of its business logic and technical behaviour. Additionally, conditionally turning certain mechanism off makes testing the code very handy.
Let's measure the effect in LOC (Line of Code) and LOI (Level of Indent) by an example of cancelling user subscription on server-side with some minimal error handling of retry and restore. The simplification effect will be magnified with increasing complexity of the control mechanism.
Using @opbi/hooks Hooks: LOC = 16, LOI = 2
// import userProfileApi from './api/user-profile'; // import subscriptionApi from './api/subscription'; // import restoreSubscription from './restore-subscription' import { errorRetry, errorHandler, chain } from '@opbi/hooks'; const retryOnTimeoutError = errorRetry({ condition: e => e.type === 'TimeoutError' }); const restoreOnServerError = errorHandler({ condition: e => e.code > 500, handler: (e, p, m, c) => restoreSubscription(p, m, c), }); const cancelSubscription = async ({ userId }, meta, context) => { const { subscriptionId } = await chain( retryOnTimeoutError )(userProfileApi.getSubscription)( { userId }, meta, context ); await chain( errorRetry(), restoreOnServerError, )(subscriptionApi.cancel)({ subscriptionId }, meta, context); }; // export default cancelSubscription;
Vanilla JavaScript: LOC = 32, LOI = 4
// import userProfileApi from './api/user-profile'; // import subscriptionApi from './api/subscription'; // import restoreSubscription from './restore-subscription' const cancelSubscription = async ({ userId }, meta, context) => { let subscriptionId; try { const result = await userProfileApi.getSubscription({ userId }, meta, context); subscriptionId = result.subscriptionId; } catch (e) { if(e.type === 'TimeoutError'){ const result = await userProfileApi.getSubscription({ userId }, meta, context); subscriptionId = result.subscriptionId; } throw e; } try { try { await subscriptionApi.cancel({ subscriptionId }, meta, context); } catch (e) { if(e.code > 500) { await restoreSubscription({ subscriptionId }, meta, context); } throw e; } } catch (e) { try { return await subscriptionApi.cancel({ subscriptionId }, meta, context); } catch (e) { if(e.code > 500) { await restoreSubscription({ subscriptionId }, meta, context); } throw e; } } }; // export default cancelSubscription;
How to Use
Install
yarn add @opbi/hooks
Standard Function
Standardisation of function signature is powerful that it creates predictable value flows throughout the functions and hooks chain, making functions more friendly to meta-programming. Moreover, it is also now a best-practice to use object destruct assign for key named parameters.
Via exploration and the development of hooks, we set a function signature standard to define the order of different kinds of variables as expected and we call it action function
:
/** * The standard function signature. * @param {object} param - parameters input to the function * @param {object} meta - metadata tagged for function observability(logger, metrics), e.g. requestId * @param {object} context - contextual callable instances or unrecorded metadata, e.g. logger, req */ function (param, meta, context) {}
Config the Hooks
All the hooks in @opbi/hooks are configurable with possible default settings.
In theexample, errorRetry() is using its default settings, while restoreOnServerError is configured errorHandler . Descriptive names of hook configurations help to make the behaviour very self-explanatory. Patterns composed of configured hooks can certainly be reused.
const restoreOnServerError = errorHandler({ condition: e => e.code > 500, handler: (e, p, m, c) => restoreSubscription(p, m, c), });
Chain the Hooks
"The order of the hooks in the chain matters."
Under the hood, the hooks are implemented in the decorators pattern. The pre-hooks, action function, after-hooks/error-hooks are invoked in a pattern as illustrated above. In theexample, as errorRetry(), restoreOnServerError are all error hooks, restoreOnServerError will be invoked first before errorRetry is invoked.
Ecosystem
Currently available hooks:
Hooks are named in a convention to reveal where and how it works [hook point][what it is/does]
, e.g. errorCounter, eventLogger . Hook points are named before, after, error
and event
(multiple points).
Extension
You can easily create more standardised hooks with addHooks helper. Open source them aligning with the above standards via pull requests or individual packages are highly encouraged.
Decorators
Hooks here are essentially configurable decorators, while different in the way of usage. We found the name 'hooks' better describe the motion that they are attached to functions not modifying their original data process flow (keep it pure). Decorators are coupled with class methods, while hooks help to decouple definition and control, attaching to any function on demand.
//decorators class SubscriptionAPI: //... @errorRetry() cancel: () => {}
//hooks chain( errorRetry() )(subscriptionApi.cancel)
Adaptors
To make plugging in @opbi/hooks hooks to existing systems easier, adaptors are introduced to bridge different function signature standards.
const handler = chain( adaptorExpress(), errorRetry() )(subscriptionApi.cancel) handler(req, res, next);
Refactor
To help adopting the hooks by testing them out with minimal refactor on non-standard signature functions, there's an unreleased adaptor to bridge the function signatures. It is not recommended to use this for anything but trying the hooks out, especially observability hooks are not utilised this way.
Reducers
Integration with Redux is TBC.
Pipe Operator
We are excited to see how pipe operator will be rolled out and hooks can be elegantly plugged in.
const cancelSubscription = ({ userId }, meta, context) |> chain(timeoutErrorRetry)(userProfileApi.getSubscription) |> chain(restoreOnServerError, timeoutErrorRetry)(subscriptionApi.cancel);
Inspiration
License
以上所述就是小编给大家介绍的《Automate error handling and observarability with hooks》,希望对大家有所帮助,如果大家有任何疑问请给我留言,小编会及时回复大家的。在此也非常感谢大家对 码农网 的支持!
猜你喜欢:本站部分资源来源于网络,本站转载出于传递更多信息之目的,版权归原作者或者来源机构所有,如转载稿涉及版权问题,请联系我们。
500 Lines or Less
Amy Brown、Michael DiBernardo / 2016-6-28 / USD 35.00
This book provides you with the chance to study how 26 experienced programmers think when they are building something new. The programs you will read about in this book were all written from scratch t......一起来看看 《500 Lines or Less》 这本书的介绍吧!