投稿時間:2021-05-16 10:16:54 RSSフィード2021-05-16 10:00 分まとめ(17件)

カテゴリー等	サイト名等	記事タイトル・トレンドワード等	リンクURL	頻出ワード・要約等/検索ボリューム	登録日
python	Pythonタグが付けられた新着投稿 - Qiita	scikit-learn付属のデータセットをいつでも使えるように準備しておく	https://qiita.com/azumabashi/items/0afd57f7f789f8caa31a	情報元公式環境pythonsklearnnumpyscipyCythonpandasデータセットの種類データセット取得方法用途ボストン市の住宅価格データloadbostonreturnXy回帰アヤメの品種データloadirisreturnXyasframe分類糖尿病患者の診療データloaddiabetesreturnXyasframe回帰数字の手書き文字データloaddigitsnclassreturnXyasframe分類生理学的特徴と運動能力の関係についてのデータloadlinnerudreturnXyasframeワインの品質データloadwinereturnXyasframe分類乳がんデータloadbreastcancerreturnXyasframe分類ボストン市の住宅価格データBostonhousepricesdatasetUCIML住宅データセットのコピーです。	2021-05-16 09:01:16
js	JavaScriptタグが付けられた新着投稿 - Qiita	【Javascript】 分割代入について（オブジェクト編）ーDestructuring Objects	https://qiita.com/redrabbit1104/items/c8eb9d88e238ec0b59fd	オブジェクトはkeyとvalueのつの要素になっているので、変数名はjavascript側で認識できるような形にしなければなりません。	2021-05-16 09:34:40
Program	[全てのタグ]の新着質問一覧｜teratail（テラテイル）	プライベートサブネットにあるEC2インスタンスにアクセスできない	https://teratail.com/questions/338538?rss=all		2021-05-16 09:28:48
Program	[全てのタグ]の新着質問一覧｜teratail（テラテイル）	checkboxのチェックをデータベースから反映させたいのですができません	https://teratail.com/questions/338537?rss=all	checkboxのチェックをデータベースから反映させたいのですができませんkenkenpythoncomこのページの一番最後を見ているのですが、関数ベースのviewしかのってなくて、クラスベースで作りたいのですがうまくいきません。	2021-05-16 09:24:02
Ruby	Rubyタグが付けられた新着投稿 - Qiita	Railsポートフォリオ #15 READMEの作成	https://qiita.com/yanoo/items/eff554fe284b2f2266cd	前回記事カテゴリーソート機能ポートフォリオの顔とも言える超重要なReadme今回はポートフォリオにおいて最も重要なものの一つであるReadmeの作成に取り組みました。	2021-05-16 09:35:00
Ruby	Rubyタグが付けられた新着投稿 - Qiita	Rubyで解くAtCoder ABC 201 (C問題 Rubyでは最短実行時間)	https://qiita.com/AmaiOmikan/items/f8b1d48c7e7d05627aee	高校数学でこんなのやったなあと思い出しながらoが個以上桁の暗証番号なのでこれはありえない通りoが個個全てをパターンしかないので通りoが個oの種類のみで考えられるパターンに対し同じ数字が個あるのでで割ってどの数字を個使うかでパターンあるのでをかけます。	2021-05-16 09:20:38
AWS	AWSタグが付けられた新着投稿 - Qiita	AWS Organizationを使った環境別アカウントの作成とスイッチロールの導入	https://qiita.com/morita-toyscreation/items/eeba120471cfad628c49		2021-05-16 09:10:53
AWS	AWSタグが付けられた新着投稿 - Qiita	goofysでAmazon S3をLinuxサーバにマウントする	https://qiita.com/latin1/items/82140eb6de8f7aa368fe	benchmark環境・新規構築したAmazonLinuxを使用・ECのIAMロールにSFullAccessを付与している設定手順goofysのインストール最初にgoとfuseをインストールします。	2021-05-16 09:00:55
Ruby	Railsタグが付けられた新着投稿 - Qiita	Railsポートフォリオ #15 READMEの作成	https://qiita.com/yanoo/items/eff554fe284b2f2266cd	前回記事カテゴリーソート機能ポートフォリオの顔とも言える超重要なReadme今回はポートフォリオにおいて最も重要なものの一つであるReadmeの作成に取り組みました。	2021-05-16 09:35:00
海外TECH	DEV Community	Configure Your Laravel Queues with AWS SQS	https://dev.to/techdurjoy/configure-your-laravel-queues-with-aws-sqs-7mc	Configure Your Laravel Queues with AWS SQSIntroductionRecently I had to migrate my queue driver on a project from Redis to SQS and for some weird reason the information provided on the official Laravel Queues documentation didn t do justice to this and I couldn t find an article online that could help that s why I m writing this with the hope that you don t spend two hours of your Saturday morning figuring stuff that should have been documented I d try to keep this article as brief as possible I d attach links to other articles that explain some steps so I can focus on the important parts PrerequisitesAn existing Laravel applicationAn AWS accountAn understanding of Laravel queuesSome AWS knowledgeIf you use Laravel Horizon to manage your queues you d be making some modifications to your horizon php file For each worker defined change the queue connection from redis which is the default to env QUEUE CONNECTION Out of the box Laravel dispatches jobs to the default queue therefore your first queue should be named default on SQS Configure Your Laravel Queues with AWS SQS	2021-05-16 00:33:59
海外TECH	DEV Community	Honeycomb, why are my service's requests so slow?!	https://dev.to/shayde/honeycomb-in-action-why-are-my-service-s-requests-so-slow-516g	Honeycomb why are my service x s requests so slow Honeycomb can be used to help identify the best bang for your buck in terms of time spent optimizing for performance Suppose I have been asked to spend the next sprints researching and contributing code changes and improvements to help increase service performance The only information I have been given from customers is Things run slowly right after I login during certain hours of the day Needle haystack right Okay at least I have some starting points Honeycomb should still allow me to take this broad symptom of slow requests and investigate further Trace data includes timestamps and duration fields which should allow me to get some information about the requests my system sees over time Our customers seem to notice this most directly after signing in I know that after signing in and anytime a user accesses common parts of the platform their account information is queried So I think I at least know what microservice to drill into Honeycomb can help me drill into even more granular data with a simple query I construct a query for all of the events my service emits for each API request I then group them by the endpoint name and the status code result and create a visualization of the th percentile of request duration This query tells me that the GET profile endpoint responds to of requests in ms or less Knowing that this is a heavily used endpoint and compared to the GET connectioninfo endpoint that sees similar traffic but at a x faster duration it is apparent that this route is pretty slow compared to other requests to my system Further querying tells me the average response duration for the profile endpoint over that same course of time was ms Additionally P duration ms half of all requests was ms This endpoint seems like it may be a good candidate for further inspection it looks like it is used heavily and that kind of response rate could definitely be a source a slowness for end users Hmm How can I learn more What specifically about that endpoint is so slow Do I need to set up some local performance testing to figure it out Of course not Honeycomb has all of the info I need in the trace data graphs for the requests I am curious about From the query results view I can drill further into the data by hovering over the name column s value I am interested in GET profile clicking the menu that appears and selecting the option to show only results where name GET profile and http status code This is where I can put my detective hat on and start doing the interrogating I talked about in my initial post now that I have specific data and a good question to ask what is causing the most heavily endpoint of my service profile to be so damn slow I can use Honeycomb to start looking for patterns and the source of the slowness From my query above I can see that the pattern is pretty typical during the weekdays I narrow my search to hours of the work week by highlighting that time range and clicking the button that appears to zoom in my search window Oh okay at this level I can see that the data is pretty normal aside from a few outliers during off peak times I m mostly curious about peak usage and typical requests so I m going to zoom in even further to the peak usage time of that endpoint over a hour duration Awesome that s much more manageable and more typical of the usage scenario I m concerned with Once I have drilled in enough and crafted the query for the request data I am interested in examining I am going to start poking at the trace data for these requests and see if I can t find some common patterns I can access trace data for requests a few different ways The Traces tab in Honeycomb will give me the slowest traces that Honeycomb has for the time range of my query I can also head on over to the Raw Data tab and click on the trace trace id column value for any event I am curious about Clicking into a trace shows me a time series waterfall visualization of the request The trace above shows a typical request to the profile endpoint and helps me understand potential areas for improvement Keep in mind I have not even touched the code that runs this service and I have already identified key areas of it that may be in need of attention If I analyze the trace and walk through the codebase alongside it I can get context about the external requests the original one generated these are notated in blue on the screenshot above Armed with the request data and its context in code I can quickly discern what the slow up is It turns out the profile endpoint is very popular in our platform it can be called dozens of times in quick succession as a user navigates to different areas of our applications This in and of itself is not problematic however every time the user s profile endpoint is called I can see in code and tell from the trace that data changed with a low frequency user first last name email avatar image url user permissions and rights to products on the platform etc is being retrieved from various downstream services and ultimately resulting in multiple database executions When the same data is being queried by and for the same user multiple times in quick succession this results in more resource usage by the server slower response times and greater database traffic spend especially in PaaS database models that charge on a request transaction throughput basis The highest trafficked endpoint in our service also happens to be reading the same rarely changed data from a database for each user oftentimes in rapid succession several times in seconds This is a ripe candidate for a smart in memory caching mechanism Armed with the request duration data I am now able to tell a full story and pinpoint a worthwhile time to spend investing in performance for my service all without doing any sort of local performance testing or even looking at the logic in the code Allow the observability data to give you the data and answers you are looking for It can help identify specific the microservices endpoints and areas of the codebase to look at so you are ultimately making the best use of your time and efforts especially in companies and enterprises with hundreds of microservices supporting a large platform In my next post I ll show how Honeycomb and observability trace data can be useful tools to help implement smart caching mechanisms to balance service performance and cost I ll also highlight some other features in Honeycomb such as Boards and Derived Columns stay tuned Learn more straight from the Hive Explore trace dataView your raw dataCreating and using Heatmaps	2021-05-16 00:07:42
ニュース	BBC News - Home	Israel Gaza conflict: Netanyahu vows to continue strikes	https://www.bbc.co.uk/news/world-middle-east-57131272	attacks	2021-05-16 00:42:40
LifeHuck	ライフハッカー［日本版］	長距離ランナーのための、黒くなった爪の適切な対処法	https://www.lifehacker.jp/2021/05/what-to-do-when-running-gives-you-black-toenails.html	長距離	2021-05-16 10:00:00
北海道	北海道新聞	広範囲で雷伴う激しい雨の恐れ　梅雨前線、低気圧が影響	https://www.hokkaido-np.co.jp/article/544340/	梅雨前線	2021-05-16 09:17:00
北海道	北海道新聞	米大統領、市民の犠牲懸念　パレスチナ政府議長と初電話会談	https://www.hokkaido-np.co.jp/article/544339/	米大統領	2021-05-16 09:17:00
北海道	北海道新聞	小平２５位に浮上、松山は４４位　米男子ゴルフ第３日	https://www.hokkaido-np.co.jp/article/544337/	男子ゴルフ	2021-05-16 09:10:00
北海道	北海道新聞	高齢者向け７月中接種「８５％」に懐疑的　政府調査に道内自治体　医師ら確保の担保なく	https://www.hokkaido-np.co.jp/article/544304/	新型コロナウイルス	2021-05-16 09:08:06