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。

How we moved to Google Cloud using Consul and ZeroTier with zero downtime

105 点作者 rkrzr超过 7 年前

5 条评论

vdloo超过 7 年前
This is very cool. I&#x27;m wondering if decentralized overlay &#x27;mesh&#x27; networks will become more prevalent in the future. Overlays have obvious benefits for multi-cloud setups if latency isn&#x27;t an issue for the services you run through the network. I can imagine this becoming a more popular technique in the future as cloud instances become cheaper and more people are willing to make the performance trade-off for convenience. Additionally it can be a great pattern to fight vendor lock-in.<p>I run a similar concoction for my &#x27;home network&#x27; which consists of various mobile devices like my Android phone and some cloud servers, but instead of ZeroTier and I use CJDNS in combination with Consul. The code for it is on github in the vdloo&#x2F;raptiformica repo. I ran into various issues with the difference in latency between nodes. I think most people run Consul in a very homogeneous environment (like in one datacenter), but maybe perhaps the differences between using it cross-cloud is not enough to cause problems. I&#x27;m wondering if there were some Consul settings that the author had to tweak (and how) for stability and if there were any unexpected issues.<p>One thing that caused me problems with Consul on overlay networks was an ARP cache overflow. DigitalOcean also ran into that running Consul at scale in their DC if I recall correctly: <a href="http:&#x2F;&#x2F;youtu.be&#x2F;LUgE-sM5L4A" rel="nofollow">http:&#x2F;&#x2F;youtu.be&#x2F;LUgE-sM5L4A</a> I noticed that if I put enough Dockers on one host (like 50 - 100) in an overlay network and tried to run Consul on top of that things would start to overflow, presumably because of all the network abstraction and redundancy. I&#x27;m wondering how many machines the author had in one Consul cluster and if they tested to what amount of nodes this setup could scale.
freedomben超过 7 年前
This is a fantastic write-up. Also I never knew this about S3:<p>&quot;While it does provide read-after-write consistency for new files, it only provides eventual consistency for overwrite PUTs and for DELETEs.&quot;
评论 #15548698 未加载
评论 #15552699 未加载
评论 #15551881 未加载
williamstein超过 7 年前
&gt; &quot;With GCP, the pricing is very straightforward: you pay a certain base price per instance per month, and you get a discount on that price of up to 80% for sustained use. In practice: If you run an instance 24&#x2F;7, you end up paying less than half of the “regular” price.&quot;<p>I wish either of those two statements were true! I hope you didn&#x27;t actually base your pricing decisions on what you wrote in the article. The sustained use discount is a total of 30% off (not &quot;more than 50%&quot;), if you use an instance 24&#x2F;7 for a calendar month. Also, the 80% discount off the full price is only for pre-emptible instances, which are the ones that may not be available and are always killed within 24 hours.
评论 #15552404 未加载
api超过 7 年前
Founder of ZeroTier here -- nice writeup!
评论 #15552689 未加载
评论 #15553536 未加载
评论 #15552859 未加载
alexandre_m超过 7 年前
I personally would try wireguard instead of Zerotier for those hybrid cloud topologies instead.<p>Of course it&#x27;s a kernel module, so you have to build for target host and load.<p>Anyone tried this?<p>Been using it for internal apps (to mesh kubernetes nodes in virtual deployments) successfully for some months.