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。

AWS Application Load Balancer

371 点作者 rjsamson将近 9 年前

30 条评论

encoderer将近 9 年前
We plan to do a blog post about this at some point, but we had the pleasure of seeing exactly how elastic the elb is when we switched Cronitor from linode to aws in February 2015. Requisite backstory: Our api traffic comes from jobs, daemons, etc, which tend to create huge hot spots at tops of each minute, quarter hour, hour and midnight of popular tz offsets like UTC, us eastern, etc. There is an emergent behavior to stacking these up and we hit peak traffic many many times our resting baseline. At the time, our median ping traffic was around 8 requests per second, with peaks around 25x that.<p>What&#x27;s unfortunate is that in the first day after setting up the elb we didn&#x27;t have problems, but soon after we started getting reports of intermittent downtime. On our end our metrics looked clean. The elb queue never backed up seriously according to cloud watch. But when we started running our own healthchecks against the elb we saw what our customers had been reporting: in the crush of traffic at the top of the hour connections to the elb were rejected despite the metrics never indicating a problem.<p>Once we saw the problem ourselves it seemed easy to understand. Amazon is provisioning that load balancer <i>elastically</i> and our traffic was more power law than normal distribution. We didn&#x27;t have high enough baseline traffic to earn enough resources to service peak load. So, cautionary tale of dont just trust the instruments in the tin when it comes to cloud iaas -- you need your own. It&#x27;s understandable that we ran into a product limitation, but unfortunate that we were not given enough visibility to see the obvious problem without our own testing rig.
评论 #12269555 未加载
评论 #12269510 未加载
评论 #12271337 未加载
评论 #12273357 未加载
评论 #12272811 未加载
评论 #12272785 未加载
ihsw将近 9 年前
Can we agree on the terminology for Application Load Balancer and Elastic Load Balancer?<p>* ALB: Application Load Balancer<p>* ELB: Elastic Load Balancer<p>I have seen Application Elastic Load Balancer&#x2F;AELB, Classic Load Balancer&#x2F;CLB, Elastic Load Balancer (Classic)&#x2F;ELBC, Elastic Load Balancer (Application)&#x2F;ELBA.<p>In any event, I think it is great that AWS is bringing WebSockets and HTTP&#x2F;2 to the forefront of web technology.
评论 #12271724 未加载
tobz将近 9 年前
The real question: does this provide a faster elasticity component than ELBs?<p>At a previous employer, we punted on ever using ELBs at the edge because our traffic was just too unpredictable.<p>Combining together all of the internet rumors, I&#x27;ve been led to believe that ELBs were&#x2F;are custom software running on simple EC2 instances in an ASG or something, hence being relatively slow to respond to traffic spikes.<p>Given that ALBs are metered, it seems like this suggests shared infrastructure (binpacking peoples ALBs onto beefy machines) which makes me wonder if that is how it actually works now, because it would seem the region&#x2F;AZ-level elasticity of ALBs could actually help the elasticity of a single ALB.<p>If you don&#x27;t have to spin up a brand new machine, but simply configure another to start helping out, or spin up a container on another which launches faster than an EC2 instance... that&#x27;d be clutch.<p>Deep thoughts?
0xmohit将近 9 年前
AWS still doesn&#x27;t support IPv6. Good to see them talking about HTTP&#x2F;2.<p>Waiting for AWS to embrace IPv6.
评论 #12268909 未加载
评论 #12268954 未加载
评论 #12270535 未加载
评论 #12270435 未加载
评论 #12268918 未加载
boundlessdreamz将近 9 年前
So this is pretty much the same as Google HTTP load balancing <a href="https:&#x2F;&#x2F;cloud.google.com&#x2F;compute&#x2F;docs&#x2F;load-balancing&#x2F;http&#x2F;" rel="nofollow">https:&#x2F;&#x2F;cloud.google.com&#x2F;compute&#x2F;docs&#x2F;load-balancing&#x2F;http&#x2F;</a> + websocket &amp; http2?
评论 #12269745 未加载
评论 #12272076 未加载
fred256将近 9 年前
+1 for CloudFormation support on launch day. +1 for support for ECS services with dynamic ports (finally!) -1 for no CloudFormation support for ECS<p>(To configure an ECS service to use an ALB, you need to set a Target Group ARN in the ECS service, which is not exposed by CloudFormation)
评论 #12328969 未加载
评论 #12271327 未加载
评论 #12271328 未加载
cheald将近 9 年前
Exciting! Disappointing that you can&#x27;t route based on hostname yet, though. I&#x27;ve got 5 ELBs set up to route to different microservices for one app, and because we couldn&#x27;t do path-based routing before, that&#x27;s all segmented by hostname. As soon as ALB supports hostname routing, I can collapse those all into a single LB.
评论 #12269475 未加载
agwa将近 9 年前
&gt; 25 connections&#x2F;second with a 2 KB certificate, 3,000 active connections, and 2.22 Mbps of data transfer or<p>&gt;5 connections&#x2F;second with a 4 KB certificate, 3,000 active connective, and 2.22 Mbps of data transfer.<p>&quot;2KB certificate&quot; and &quot;4KB certificate&quot;? Is this supposed to read &quot;2048 bit RSA&quot; and &quot;4096 bit RSA&quot;?
评论 #12269042 未加载
indale将近 9 年前
This looks pretty sweet. The next big thing for api versioning would be header instead of url based routing, looking forward to &#x27;give you access to other routing methods&#x27;.
评论 #12269424 未加载
rjsamson将近 9 年前
They finally added support for websockets! Really looking forward to giving this a try with Phoenix.
评论 #12268844 未加载
评论 #12270165 未加载
daigoba66将近 9 年前
These new features are cool... but they still pale in comparison to something like HAProxy.<p>I guess the tradeoff is that with ELB&#x2F;ALB, like most PaaS, you don&#x27;t have to &quot;manage&quot; your load balancer hosts. And it&#x27;s probably cheaper than running an HAProxy cluster on EC2.<p>But for the power you get with HAProxy, is it worth it?<p>Does anyone have experience running HAProxy on EC2 at large scale?
评论 #12272309 未加载
erikcw将近 9 年前
I&#x27;m curious if this will Convox to route to multiple services with just a single ALB instead of the historical default of 1 ELB per service. Would be a real cost savings for a micro-services architecture.
评论 #12269571 未加载
avitzurel将近 9 年前
This is very good. Recently my workflow has been ELB -&gt; NGINX -&gt; Cluster.<p>Nginx was a cluster of machines that did routing based on rules into the ec2 machines. Now that the AELB has some of those capabilities it&#x27;s time to evaluate it.
archgrove将近 9 年前
Any love for Elastic Beanstalk with these? They seem well matched. Though EB always feels a bit of a red-headed stepchild in the AWS portfolio.
评论 #12269308 未加载
dblooman将近 9 年前
It seems that routing is done in the following way &#x2F;API&#x2F;* goes to applications and expects :8080&#x2F;api&#x2F; rather than the root. Would be nice to have the option to direct traffic to just :8080.
评论 #12271564 未加载
axelfontaine将近 9 年前
It looks like the big missing piece is auto-scaling groups as target groups...
评论 #12272992 未加载
评论 #12271456 未加载
sturgill将近 9 年前
This sentence sums up one of my main reasons for appreciating AWS:<p>The hourly rate for the use of an Application Load Balancer is 10% lower than the cost of a Classic Load Balancer.<p>They frequently introduce new features while cutting costs.
评论 #12269453 未加载
shawn-butler将近 9 年前
Anybody know whether the new ALB handles a client TLS (SSL) when operating in http mode?<p>I was trying secure an API Gateway backend using a client certificate but found ELB doesn&#x27;t currently support client side certificates when operating in http mode.<p>There was this complicated Lambda proxy workaround solution but I gave up halfway through...<p><a href="https:&#x2F;&#x2F;aws.amazon.com&#x2F;blogs&#x2F;compute&#x2F;using-api-gateway-with-vpc-endpoints-via-aws-lambda&#x2F;" rel="nofollow">https:&#x2F;&#x2F;aws.amazon.com&#x2F;blogs&#x2F;compute&#x2F;using-api-gateway-with-...</a>
kookster将近 9 年前
As a heavy ECS user, all I can say is thank you, finally!
评论 #12269239 未加载
renaudg将近 9 年前
I&#x27;m the process of containerizing an app that includes a Websockets service, and given ECS &#x2F; ELB limitations we&#x27;d just decided to go for Kubernetes as the orchestration layer.<p>This ALB announcement + the nicer ECS integration could tip the balance though.<p>Any thoughts on how likely it is that Kubernetes can&#x2F;will take advantage of ALBs (as Ingress objects I suppose) soon ?
nodesocket将近 9 年前
Do ALBs support more than a single SSL certificate?
评论 #12269874 未加载
评论 #12269113 未加载
评论 #12269324 未加载
manishsharan将近 9 年前
This is definitely nicer than having to create subdomains for microservices and mapping each subdomain url to its own Elastic Loaad Balancer + Elastic Beanstalk instance. But I have already gone down this path so I am unlikely to use AWS Application Load balancer. I wish I had this option a year ago.
nailer将近 9 年前
Nice haproxy &#x2F; nginx alternative. It&#x27;s got http2 support though which puts it ahead of haproxy.
评论 #12269945 未加载
DonFizachi将近 9 年前
Any idea if sticky TCP sessions will be supported on ELB&#x2F;ALB any time soon?
amasad将近 9 年前
I wonder if they fixed the routing algorithm for TCP connections. It&#x27;s round-robin on ELB, which is performs terribly for long lasting connections.
nodesocket将近 9 年前
So what would be a use case for using ELBs now? Seems like ALBs do everything ELBs do, but with websocket and HTTP&#x2F;2 support.
评论 #12268990 未加载
评论 #12269191 未加载
评论 #12269946 未加载
评论 #12269207 未加载
joneholland将近 9 年前
Disappointing. I was hoping they were launching a service discovery stack to compliment ECS.
bradavogel将近 9 年前
Does anyone know if it (finally) supports sticky websocket sessions?
评论 #12270505 未加载
评论 #12270172 未加载
merb将近 9 年前
Virtual Host Load Balancer would be great.
NeckBeardPrince将近 9 年前
Any idea if it&#x27;s HIPPA compliant?
评论 #12284139 未加载