﻿<?xml version="1.0" encoding="utf-8"?><rss version="2.0"><channel><title>Ayende @ Rahien</title><link>http://ayende.com</link><description>Ayende @ Rahien</description><copyright>Copyright (C) Ayende Rahien  2004 - 2021 (c) 2026</copyright><ttl>60</ttl><item><title>Christopher Wright commented on RavenHQ &amp;amp; Amazon EC2 Outage</title><description>@Ayende: Ah, okay, that's a lot better than I initially thought. I should have remembered that you're smart.

I personally would never trust anything important to just two zones, but I understand that a lot of companies are budget-conscious and can take a short outage, especially if they can point at someone else to blame.</description><link>http://ayende.com/156737/ravenhq-amazon-ec2-outage#comment6</link><guid>http://ayende.com/156737/ravenhq-amazon-ec2-outage#comment6</guid><pubDate>Mon, 02 Jul 2012 16:47:24 GMT</pubDate></item><item><title>Colin Bull commented on RavenHQ &amp;amp; Amazon EC2 Outage</title><description>Hummm, wondering whether this was the pesky leap second introduced on 30 June at midnight..</description><link>http://ayende.com/156737/ravenhq-amazon-ec2-outage#comment5</link><guid>http://ayende.com/156737/ravenhq-amazon-ec2-outage#comment5</guid><pubDate>Mon, 02 Jul 2012 14:48:33 GMT</pubDate></item><item><title>Ayende Rahien commented on RavenHQ &amp;amp; Amazon EC2 Outage</title><description>Christopher,
We are running in multiple availability zones, and we are putting non replicated clients in multiple zones.
I'll post a full discussion of this tomorrow, but the reason you saw a lot of activity about that is that most of our free plans resided on that region.

"The secondary node" is actually a RavenDB term, which refers to the secondary node that you fail to.
The question if you have  tertiary or more actually depend on your plan. And N node in a replicated plan are on a different availability zone.</description><link>http://ayende.com/156737/ravenhq-amazon-ec2-outage#comment4</link><guid>http://ayende.com/156737/ravenhq-amazon-ec2-outage#comment4</guid><pubDate>Sun, 01 Jul 2012 19:39:12 GMT</pubDate></item><item><title>Christopher Wright commented on RavenHQ &amp;amp; Amazon EC2 Outage</title><description>Have you considered hosting RavenHQ servers in multiple availability zones and regions, and randomizing your non-replicated customers between zones?

It sucks to have a large portion of your customer base experience an outage because of you. By hosting in multiple zones, you can reduce the portion of your customer base that any single incident affects.

Also, when you say "the secondary node", I immediately get worried. Is there a tertiary node? Are all the secondary nodes in the same zone as each other?</description><link>http://ayende.com/156737/ravenhq-amazon-ec2-outage#comment3</link><guid>http://ayende.com/156737/ravenhq-amazon-ec2-outage#comment3</guid><pubDate>Sat, 30 Jun 2012 16:44:00 GMT</pubDate></item><item><title>Edward Spelt commented on RavenHQ &amp;amp; Amazon EC2 Outage</title><description>Netflix, Pinterest en Instagram are/where offline too.</description><link>http://ayende.com/156737/ravenhq-amazon-ec2-outage#comment2</link><guid>http://ayende.com/156737/ravenhq-amazon-ec2-outage#comment2</guid><pubDate>Sat, 30 Jun 2012 11:01:44 GMT</pubDate></item><item><title>James Manning commented on RavenHQ &amp;amp; Amazon EC2 Outage</title><description>Needs more chaos monkey!

http://www.codinghorror.com/blog/2011/04/working-with-the-chaos-monkey.html

Fail early, fail often! :)</description><link>http://ayende.com/156737/ravenhq-amazon-ec2-outage#comment1</link><guid>http://ayende.com/156737/ravenhq-amazon-ec2-outage#comment1</guid><pubDate>Fri, 29 Jun 2012 16:30:54 GMT</pubDate></item></channel></rss>