Is the cloud safe for my data?

There are a lot of different sides to this question.

One aspect might be, are my servers and what they contain as safe in the cloud as my own in-house network? Amazon Web Services supports a VPC setup, allowing you to essentially extend your corporate network, and continue to use the same perimeter security and firewall you’re used to.

What about endpoint security of individual servers? There is no reason that endpoint security supported by Amazon security groups must be thought of as any more or less secure than perimeter security. In fact you can simulate perimeter security of a firewall by using a database group, a webserver group, and a jump box for ssh.

What about my data itself? Data residing on the server itself is quite secure, given the above considerations. However snapshots and backups of your data will be stored in S3. As it is a new technology, you may want to consider how secure it is and whether vulnerabilities may be found in the future. For particularly sensitive data you may choose to either encrypt the filesystem, or a volume where sensitive data itself resides. Or you might choose to store very sensitive data outside the cloud.

How do I make this part of my website load faster?

There are many moving parts in the plumbing of the internet. There is data in transit from your end users to your hosted servers, there is network latency within your hosting environment, web server and caching layer challenges, and then the backend database tier. Your pipes can get clogged in any of those areas – we call them performance bottlenecks.

Troubleshooting performance and scalability involves investigating where the source of the trouble is, often a bit of detective work. Once the source is isolated, we identify the cause, and mitigate the problem, adding new pipes to scale as necessary.

We lost our data and restored with our backup but some things are missing. Why?

Running regular fire drills on your backups is the only way you can be sure they are complete. Data could be missing, you could be backing up the wrong sets of data, or to the wrong point in time. Furthermore you could be missing some configuration or code when restoring, leaving one component still awry.

Can you be present on site throughout the project?

For larger engagements, we typically spend time on site at the start. Once we’ve worked closely with the team, and built a comfortable working relationship, work is typically done off-site. Of course all of this is on a case by case basis, as we find each client’s needs are different.

I’m thinking of migrating to the cloud. Which service is better?

There are many to choose from, including Rackspace, Amazon, Logicworks, Softlayer, Datapipe, Joyent and more. Although in certain respects Amazon is furthest along in terms of flexibility, Rackspace provides better service. So there are pros and cons to each and we can help you evaluate the best for your needs.

IaaS, PaaS, DBaaS. What’s the difference?

IaaS means infrastructure, and all the Amazon EC2 instances we talk about in our articles and content. This is basically raw iron.

PaaS means platform, such as Heroku, EngineYard or DotCloud and so forth. Although these hosting solutions are said to promote easy scalability, they can struggle with plumbing problems too. We can troubleshoot performance here as well.
DBaaS means Database as a service. You could be using a solution such as Xeround or Amazon’s own RDS. In either case we can help you with issues and challenges.

Our technical team are located in {country}. Can you work with different time zones?

Absolutely. We work with firms as far afield as South Asia and Australia. We find regular overlapping times to work closely with your team as needs dictate.

Can you help us migrate an Oracle application to MySQL?

Migrating applications from one relational database to another can be an involved process. From stored procedures, triggers and functions, to datatype dependencies, applications are rarely built database independently. What’s more SQL code will typically need to be rewritten or optimized again to support MySQL as the optimizer in MySQL is not as sophisticated as Oracle. In any case it can be done, and building a step-by-step plan is the first move towards that goal.
Also be mindful that there is a cultural difference between the Oracle world of enterprise applications and legacy solutions, and the free-wheeling open-source world where many solutions are often do-it-yourself and roll your own.

Can you help us migrate a MySQL application to Oracle?

Certainly MySQL applications can be migrated to Oracle. As with migrating in the reverse direction, optimizer and language differences will surely challenge a casual attempt to migrate. Plan for plenty of time to test, iron out bugs, and work through issues and challenges.

Can we take baby steps to the cloud?

Absolutely. It’s the recommended way. Wade into the water, and learn to swim before diving into the deep end. There are a lot of ‘gotchas’ in cloud computing from learning new concepts, to managing virtual resources. When your team is immersed in traditional hosting center operations where physical machines, cables, and components are involved, the step to a very abstract all virtualized environment is both liberating and challenging at the same time. New performance challenges rear their ugly heads as well, from shared network and disk resources that you’re used to having all to yourselves.

Can you help bring our engineers up to speed on technologies such as MySQL or Amazon EC2?

One thing we promise is to work closely with your engineers. Starting with best practices we’ll work side by side with your team to build out scalable solutions, and at the same time help them get up to speed with open-source technologies, and the best solutions. After deployment and launch, we can continue to provide full service operations, or limited as-needed questions and answers to your team.

Can you build infrastructures that support zero downtime for us? Our application cannot afford any downtime at all.

High Availability is a core competency for us. We can build load balanced solutions to keep your webserver tier always on. Take individual servers offline to reconfigure, patch or upgrade, while the other nodes continue to server your website. The same thing can be done with the database tier using MySQL’s multi-master configuration. While one database is serving your webservers, the other node can be taken offline to reconfigure, patch or upgrade as well. You can even use this method to perform slow running table operations on the inactive node in your cluster.

How do you charge?

We outline a new contract with different options for each new prospect we pitch to. We don’t have fixed rates but rather adjust and negotiate for each project. Some are longer term engagements involving multiple resources, while others are short one-off jobs. For details visit our pricing page.