TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Ask HN: How does Google Cloud keep so many timeseries/graphs quick?

2 点作者 kevinsimper超过 3 年前
I am using Google Cloud and others and I am amazed by how many graphs there are everywhere, showcasing you all kind of metrics. And the crazy part is that you can define any time periode and it is equally fast.<p>Does somebody know how they designed that?<p>It is not bigquery, because bigquery is fast but multiple seconds to respond to queries.<p>It could be bigtable, but it just seems expensive and keeping this many timeseries up to date all the time and few of them are looked at that often. Bigtable starts at $500 usd for a single node and you need multiple.<p>I understand that there is a lot of money in Cloud. I also tried to search and could not find any resources on how this is done on large scale.<p>How can I do something like that in my own applications?<p>I have looked at Postgres TimescaleDB and other timeseries databases, Prometheus, but you quickly end up with a lot of timeseries that takes a lot of memory to compute (Prometheus running out of memory).<p>They have made an article here talking about OpenTSDB https:&#x2F;&#x2F;cloud.google.com&#x2F;architecture&#x2F;monitoring-time-series-data-opentsdb

暂无评论

暂无评论