There is nothing wrong with the management of databases in somewhat older environments. On the contrary, these are applications that keep a lot of companies afloat. But there are a number of things you need to pay attention to. Roland van Veen, OptimaData DBA and Database Reliability Engineer, explains in this blog what those things are.
OptimaData's multi-platform strategy certainly includes the management of databases in somewhat older environments. This may not seem very exciting and therefore may not attract as much attention, but that does not make the work less important. On the contrary, more than ever these are applications that keep companies afloat.
Key words are in this case not 'shiny new', 'flashy tools' and 'hot off the press' - of course, we have written enough other blogs about that - but 'experience', 'reliability' and 'solidity'. These three words do not only refer to the software environment, but also to the capabilities of the developers and administrators. But what should you pay attention to when dealing with a somewhat older environment? Roland van Veen, OptimaData DBA and Database Reliability Engineer, takes you with him.
1 Appreciate your workhorse and keep taking care of it
Don't look at your surroundings like some kind of legacy, lying somewhere in wait for death. It's your reliable workhorse that still has to - and is able to - come along and appreciate it that way. What is seen as legacy today may have a very different value tomorrow. For example, today we see a revival of relational databases, but with multi-model capabilities.
Of course, that doesn't mean you have to sit still. Keep developing your environment further, keep up to date with product changes, end-of-life data and migration paths. Start with common problems where the risk for users is high and make hidden risks visible.
2 Make the most of the opportunities
It can be very rewarding to spread a depreciation over another number of years. Check this every year again. That gives you time to get to know the next generation of software well and at the same time to react quickly to necessary adjustments to the current platform.
It has advantages if you can read and write with your environment without having to run after new knowledge all the time. Moreover, database skills are reusable if you pay some attention to them. Conserve experience and at the same time give room for self-development.
Moreover, this creates opportunities for candidate DBA consultants or DevOps-ers to develop the right competencies in the important field of 'experience', 'reliability' and 'solidity'.
Of course, such a somewhat older environment also has disadvantages.
The costs of such a platform may be higher than those of a new environment. Think of licensing costs compared to open source or other cost models such as those used in the cloud.
- Look ahead
At some point older software can no longer be adapted to new standards, causing security holes to be lurking, as well as problems with interconnections, drivers or the inability to meet new operating system requirements.
- Attracting talent
With an older, less common environment, it can become more difficult to find administrators and developers.
3 Don't stand still
Never stand still, modernise and optimise your environment, even if it is on the threshold of retirement. In doing so, pay attention to availability, continuity and security.
- In case of performance problems, you can look at sharing, partitioning or other forms of storage.
- Manage the data lifecycle. This can have a relationship with performance, for example if you have a lot of little used data online.
- Also pay attention to data quality - a hot topic in 2020; the older a database, the more of a mess oftentimes. Adaptations, workarounds and bugs accumulate and these data errors result in application errors or incorrect information.
- Provide good documentation that is easy to access: not a big pill in the drawer of a DBA consultant, but a wiki or collaboration software like Confluence.
- Evaluate all those features your ‘legacy’ database has gotten in the last ten years. Often not all of them are used. It's just like your toolbox; it's already loaded and yet you buy that new set of keys.
- Furthermore take a look at the hardware. Modern hardware often offers ssd-storage, more memory or many more cpu-cores. Make sure this also works together.
- Modernise the management. Think about automation, version control and making it suitable for DevOps.
- Handle database modifications smartly. Do for example short 'smoke tests' or SQL unit tests.
To migrate or not to migrate? That's the question
Migrating to another database platform, operating system or cloud platform is an option, but the effort required to do so depends strongly on the application. Are you working with an older environment and are you curious whether you can go on still some time or whether it's time to migrate? Please feel free to contact us! We like to think along with you.
back to blogoverview