Before the end of next month we hope to release Snapt HAProxy! We’ve come a long way since we originally started developing our professional front-end to the load balancer, and we’re excited to show it off.
Remember to join our launch email list to be offered a discount when it’s available!
Based on feedback received from a few of our Beta testers, we’ve been able to identify the features needed for SnaptHA to move forward. One of the biggest features is advanced ACL management.
An ACL is a rule used to determine what action is required based on the incoming request. They can also be used to monitor backends or block the request all together. Having said that, the combination of conditionals within the ACL can become tricky and a fool proof method of managing them is required to get the most out of your HAProxy server.
SnaptHA treats ACLs as objects; you create an ACL and attach it to any number of required frontends. HAProxy ships with a set of predefined ACL’s that can be used such as HTTP, FALSE, TRUE, METH_CONNECT and many more depending on your version of HAProxy. These predefined ACLs are also available to be used in the same fashion as your own custom one.
Creating the ACL is made simple by a step based wizard. You can opt to either create your own ACL or select from a list of ACL templates – our version of predefined ACLs that ship with SnaptHA.
Once an ACL has been created it can be attached to your configured frontends along with the actions to be taken on that ACL.
It’s as simple as that, and the good news is that SnaptHA is able to import most existing ACLs making the integration process that much easier!
It is critical to be able to see exactly what is going on at any point in time in a high availability environment, and one of the ways Snapt presents this data for HAProxy is with live graphing. This allows you to sit back and watch the throughput and the sessions your HAProxy groups are doing – as it happens!
Above is an example of the live traffic to our group “HTTP-Farm”. When monitoring or diagnosing a problem a live graph can be worth a thousand lines of logfile.
Snapt aims to make performing routine tasks as simple as possible and adding groups and servers to your HAProxy setup might just be one of them.
Groups are a simplified way of looking at the “listening” parameters in HAProxy and within each group you would find the servers that you are load balancing. The process of adding these groups and servers is made easy by using a step based approach. The first thing to do would be configuring the parameters of the group. These parameters include balancing methods, timeouts, protocol testing and more.
Each group you add should contain servers that will perform the task being load balanced. This process is also step based allowing you to quickly add servers to your groups. Like groups, servers have many options available to them; Snapt helps you out by grouping these options into 3 different categories – “Standard Options”, “Health Checks” and “Performance”.
4 simple steps to get your HAProxy load balancing server up and running.
One of the features packaged in the Snapt Framework is the “Interception Manager”. This provides plugins to the framework such as Snapt Squid or Snapt HAProxy to transparently intercept traffic.
In the Squid example, it makes it easy to pick up HTTP (port 80) traffic and cache it without interfering with user configuration, or with HAProxy you can roll out load balancing without changing IP addresses and the like. Typically, this can get confusing and complicated – so we introduced an interception manager.
Each plugin is assigned a “space” in which it can operate to intercept traffic. Each space can have subspaces. For example, each HAProxy group effectively has its own space which can have any number of ports or IP addresses to intercept traffic on and redirect it to HAProxy.
The interception manager automatically integrates with your firewall (currently it supports PF or IPTables) and intercepts the data. You can disable and enable interception spaces without losing their configuration – allowing maintenance without downtime, and you can disable the entire system without losing your setup.
You can see an example of some live HAProxy groups intercepting data below
The Snapt HAProxy front-end automatically logs throughput, sessions, queues, errors and much more to it’s reporting database enabling advanced load reporting, such as the example below.
What you are seeing here is a look at our HTTP-Farm load balance group, which has two servers (web03 and web04) inside of it. This report (the throughput report) shows a breakdown of sessions, queues and traffic. This type of report enables you to see traffic distribution across the group; find out if any servers are falling behind (with queues) and is just one example of the massive amount of data we store.
From there you can go even further down, into a server all the way down to minutes!
It is our belief that in order to have a properly balanced and reliable environment reporting is key! Remember that you can subscribe to our release list to be the first to hear about the release of Snapt!
One of the most important features we’ve added in to HAProxy with the Snapt HA frontend is a complete notification and logging system for server statuses. This is critical in a high availability environment – we provide notifications about any server status changes (UP/DOWN etc) as well as historic logging of server performance (sessions, queues, errors and more)
You can even configure a private Twitter account and have notifications tweeted to you!
We are getting very close to being able to release the first public version of Snapt HAProxy – our professional front-end to the well known open source load balancer!
If you’d like to read more about some of the features that are available check out our product page, or if you want to be notified first join our (spam free) notification list
With work on the Snapt Framework moving along, focus is being shifted toward the completion of our two launch modules; HAProxy and Squid.
Some of the features to look forward to in Snapt Squid have already been discussed and work on plugins has started. The plugins aim to bring functionality that might not be common to all installations without adding unnecessary bloat to the Snapt Squid Module. Examples include authentication, WCCP, advanced peering and more.
The Snapt HAProxy module is a feature packed front end to the HA Proxy Load Balancer. Among many other features, it includes advanced reporting, simple yet powerful control over the config file as well as status alerts. Scheduled for release with the launch of the Snapt Framework, this will give HA Proxy the frontend needed to position itself as a competitive Open Source load balancer.
We’re excited for 2010 and for the great feedback we have received so far. Remember, by subscribing to the release list you will be the first to know when Snapt is officially launched.
Snapt isn’t just a frontend to an existing HAProxy installation – it’s also designed to allow you to start from scratch! When you first load the HAProxy Snaptin you will be taken to the “Initial Configuration” screen you see below.
If we don’t detect an existing HAProxy installation, we will allow you to install a new version. This takes you to the next step in the process as seen below.
Snapt will now automatically download the latest version and install it for you. Immediately after this you can start using the frontend and load balancing your traffic!
