availability on SWiK

MySQL 5 High Availability with DRBD 8 and Heartbeat 2

Fri, 29 Aug 2008 09:56:24 -0700

"It’s 2 AM Saturday and the phone rings. You thought you were going to have a long holiday weekend. You’ve been informed by the NOC that there’s no connection to the database server, and could you come down to see what’s wrong with it. Remote access won’t save you, there’s no response at all from the server. [...]"

Raid is obsolete

Thu, 28 Aug 2008 18:05:47 -0700

In a lot of environments.

Peter gives a nice overview why you don't always need to invest in big fat redundant hardware.

We've tackled the topic last year already ..

Now I often get weird looks when I dare to mention that Raid is obsolete ..people fail to hear the "in a lot of environments"

Obviously the catch is in the second part, you won't be doing this for your small shop around the corner with just one machine. You'll only be doing this in an environment where you can work with a redundant array of inexpensive disks. Not with a server that has to sit in a remote and isolated location.

Next to that there are situations where you will be using raid, but not for redundancy, but for disk throughput.

DRBD:What is DRBD

Thu, 14 Aug 2008 08:04:52 -0700

XEN Cluster HowTo

Tue, 12 Aug 2008 11:05:07 -0700

Xen HA cluster guide describing how to build a Xen cluster with DRBD to provide shared storage (SAN) and heartbeat to provide fail over

Preparing For EC2 Persistent Storage

Mon, 11 Aug 2008 15:04:16 -0700

AWS Service Health Dashboard - Amazon S3 Availability Event: July 20, 2008

Thu, 07 Aug 2008 13:14:08 -0700

Installing DRBD on Hardy Part 2

Thu, 07 Aug 2008 13:13:10 -0700

The second entry which demonstrates how to setup HeartBeat on Hardy

DRBD:What is DRBD

Thu, 07 Aug 2008 13:13:08 -0700

How to scale writes with master-master replication in MySQL

Wed, 06 Aug 2008 06:15:50 -0700

This post is SEO bait for people trying to scale MySQL’s write capacity by writing to both servers in master-master replication. The short answer: you can’t do it. It’s impossible.

I keep hearing this line of reasoning: “if I make a MySQL replication ‘cluster’ and move half the writes to machine A and half of them to machine B, I can increase my overall write capacity.” It’s a fallacy. All writes are repeated on both machines: the writes you do on machine A are repeated via replication on machine B, and vice versa. You don’t shield either machine from any of the load.

In addition, doing this introduces a very dangerous side effect: in case of a problem, neither machine has the authoritative data. Neither machine’s data can be trusted, but neither machine’s data can be discarded either. This is a very difficult situation to recover from. Save yourself grief, work, and money. Never write to both masters.

high availability, mysql, replication, scaling

3G iPhone Availability - Updated Every Fifteen Minutes

Sat, 26 Jul 2008 18:57:27 -0700

3G iPhone Availability - Updated Every Fifteen Minutes

Fri, 25 Jul 2008 07:02:15 -0700

Configuring Xen HA with Heartbeat for SUSE Linux Enterprise Server

Fri, 25 Jul 2008 07:01:17 -0700

Learn how to make a Heartbeat cluster with SUSE Linux Enterprise by creating a SAN for high availability storage for Xen virtual machines.

3G iPhone Availability - Updated Every Fifteen Minutes

Fri, 25 Jul 2008 06:59:13 -0700

3G iPhone Availability - Updated Every Fifteen Minutes

Thu, 24 Jul 2008 10:02:01 -0700

3G iPhone Availability - Updated Every Fifteen Minutes

Thu, 24 Jul 2008 08:59:19 -0700

3G iPhone Availability - Updated Every Fifteen Minutes

Thu, 24 Jul 2008 08:59:17 -0700

3G iPhone Availability - Updated Every Fifteen Minutes

Thu, 24 Jul 2008 08:00:17 -0700

Apple Retail Store - iPhone availability at the Apple Store

Sun, 20 Jul 2008 21:47:28 -0700

Apple Retail Store - iPhone availability at the Apple Store

Thu, 17 Jul 2008 08:59:17 -0700

High Availability Hosting with DRBD :: BobCares :: Outsourced Web Hosting Support

Tue, 08 Jul 2008 19:47:54 -0700

High Availability with DRBD and Heartbeat Presentation | CB1, INC.

Fri, 20 Jun 2008 05:18:55 -0700

highly available cluster using DRBD and Heartbeat.

Asterisk High Availability Solutions - voip-info.org

Fri, 13 Jun 2008 14:15:59 -0700

Even Worse than It Appears?

Tue, 10 Jun 2008 11:21:46 -0700

Two things today.

First, thanks to Cote, Matt, Javier and others for their kind words. I am tremendously excited to be working with Hyperic. I’ve liked the company for a long time, and I’m even more impressed by the team and the strategy now that I am spending a couple of days a week here. I haven’t abandoned retirement altogether, but I have allowed it to erode a little bit because I like this opportunity so much.

Second, I want to pile on Javier’s post on the availability issues at Amazon over the past several days.

It’s worth pointing out that it had to be a pretty lousy weekend for the people responsible for running Amazon’s infrastructure. If you take a step back, the only reason that the downtime is remarkable is because it’s so rare. Nobody blinks when Twitter goes off-line. When my own favorite retailer since 1997 disappears, you notice it, because it simply never happens. I bet that things settle down quickly and we get back to the fast and reliable storefront that we’ve all come to expect.

Javier and others have touched on a likely cause of the outage: Complexity. As systems get more moving parts, they become harder to monitor and maintain. Many hope that the move to cloud computing will make things better; as you use infrastructure in the cloud, the thinking goes, you’ll be able to rely on the cloud service provider to keep it running.

As the downtime with Amazon’s storefront demonstrates, that’s a false hope. If you rely on computing services anywhere, you need to monitor them, and you need to understand how their availability affects your operations. IT shops are running more applications — JBoss, Tomcat, MySQL, home-grown software to run their businesses, along with the laundry list of proprietary and legacy applications they’ve installed over the years. These interact with one another. Every one of these software programs, and every connection among them, is a new potential source of failure.

We hardly ever abandon old systems and infrastructure. We only add new ones. Increased complexity is an irresistible force of nature, and managing it requires new techniques and new tools.

Back to the Amazon outage specifically: I’ve seen a couple of quotes in the media from people who have said, more or less, “Gee, whatever they changed that messed things up, they should have changed it during off hours.”

The fact of the matter is that there is no longer any such thing as “off hours.” For Amazon certainly, the storefront runs constantly. It may be nighttime in North America, but it’s daylight in Eastern Europe and Asia. More and more businesses — and especially those that deliver services over the Internet — simply never get to shut their computers down for maintenance. Their operations infrastructure has to take that into account.

More software running on more hardware in more places equals more complexity. At the same time, users all over the globe expect instantaneous access to data and services from anyplace, anytime. That combination means that IT professionals are staring at some pretty serious problems. The situation is even worse than it appears, though: For many businesses, as for Amazon, if the computers go down, the money stops flowing.

I’m glad to be at Hyperic because we’re working on the hard problems. Manageability of core infrastructure is the iceberg in front of most businesses, these days.

Assembla Blog | Users demand high Subversion availability, Assembla responds

Thu, 22 May 2008 12:23:03 -0700

Concept of a GlassFish Domain ... - A Thought a Day ...

Sat, 17 May 2008 12:08:04 -0700

DRBD : Système de disque miroir (raid 1) en réseau sous Linux

Fri, 16 May 2008 03:14:06 -0700

Preparing For EC2 Persistent Storage

Fri, 16 May 2008 03:14:04 -0700

Sun StorageTek Availability Suite at OpenSolaris.org

Thu, 15 May 2008 10:19:13 -0700

allows volumes and/or their snapshots, to be replicated between physically separated servers in real time, or by point-in-time, over virtually unlimited distances

JMS Service Availability in GlassFish V2

Mon, 14 Apr 2008 08:11:28 -0700

Replication is dead, long live Replication!

Tue, 08 Apr 2008 04:56:21 -0700

Brian Aker has found general agreement with his post: "The Death of Read Replication".

Arjen Lentz says "I think Brian is right...", and Frank Mash confirmed: "what Brian says about replication, caching and memcached is very true".

Just like Video killed the Radio Star it looks like maybe Memcached killed the Replication Hierarchy!

But of course, Brian and others are talking about replication for scaling reads.

In my session on PBXT next week at the conference I will be talking about how we plan to use synchronous replication to produce an HA solution for MySQL at the engine level.

I will also discuss how some flexibility in the PBXT architecture makes it possible to actually scale writes efficiently as mentioned by Arjen in his blog.

So don't miss it:

Inside the PBXT Storage Engine
10:50am - 11:50am Thursday, 04/17/2008
Ballroom G