Here’s one of those acronyms that may make you roll your eyes, but we find it a good guide to how we manage our client’s operations. We call it our SMART server management philosophy.
SMART server management:
- S tandardize
- M onitor
- A utomate
- R eview
- T est
Over the next few blog posts, I will touch on how we use each of these concepts to manage our clients systems. Adopting these methods allows our technical team to respond faster and more precisely. This allows us to restore critical services faster or even prevent disasters from happening in the first place.
Standardization in Server Management
Every day we refer potential clients to our competitors. Why? We have chosen to standardize on a narrow but popular range of technologies. As a result, we lose significant opportunities, so why standardize and not just take every job?
By standardizing on a narrow scope of popular services, we can offer in depth expertise. More importantly, standards allow us to excel in four areas of managing IT resources:
If you have something you know that works, why not use it repeatedly? By using standards with proven history, you can improve service reliability. Standard settings for Apache, PHP and other system services become starting points for optimizations. When we make system changes, we have procedures to assure the same process is followed every time we make a change.
We are reluctant to use the latest cool software or build the latest version of something just to have the bleeding edge technology. Your clients and web visitors probably don’t care what version of Apache, PHP or MySQL you are using. What they want is for your site to be there when they type that URL in the browser. They want the site to load quickly and predictably. By sticking to standards, we can deliver higher service availability.
Using standardize methods reduces issues caused by system updates. We often see updates fail because someone has modified configuration files or used third party packages. While there are certain business and technical reasons to use alternate solutions, in most case these changes lead only to problems. The vendors release updates against the standard system software. If you change this software, there are no guarantees on how the updates may impact your services.
Increased service reliability is why we try to standardize everything we do. Coming this year, we will be launching a new configuration management tool to assure all systems under management have the exact configurations we want.
You cannot scale one-off deployments cheaply or quickly. Without some common framework, every task requires significant input or consultation of client-specific documentation. This both slows deployments and increases costs.
By standardizing your builds and methods, you can quickly deploy new resources. This allows you to quickly respond to new IT challenges without having to go through a prolonged development process.
For example, we have a LAMP cluster model that mixes cloud and dedicated servers to deliver a highly scalable LAMP hosting platform. Because we use a standardize approach for all clients, we can launch this platform quickly. The solution is also flexible to scale up and down as needed. None of this would be possible if we had to go back to the drawing board with every new cluster deployment.
“Joe’s not here today. You will have to wait until he gets back.” If you don’t standardize, you become tied to very specific individuals versus and organization. With clear standard operating procedures and similarly trained staff, anyone within a group can fix most problems.
Within any group, some people are better at some tasks than others, but in most cases, urgent work is a simple issue. If you standardize staff training and procedures, then anyone on the team can address basic issues.
rackAID has dozens of procedures that are easily accessed by any member of our team. We constantly revise and update procedures as the need changes. We even try to push this consistency in how we triage and respond to tickets.
By using a standardize ticket handling method, we can consistently deliver services as a high level.
Be SMART: Standardize
We push and push to standardize operations, ticket handling and our operations. This may seem like a no-brainer but I often find small IT firms fail to adopt consistent procedures. Your experience varies greatly from one engagement to the next. You get varying answers to the same question. As a result service reliability suffers and launching new services becomes even more challenging.
Next Up: Monitoring
Monitoring is the life blood of our services. We check 100’s of items every minute across our clients systems. As a result, we know about problems before they become critical.