The team at Metasphere had noticed a trend of shrinking I.T. budgets, which resulted in some clients losing staff with the necessary skills to effectively manage the Oracle backend databases required by the Metasphere application. These clients were turning to Metasphere for extra support.
Whilst welcoming the opportunity to provide additional support services, this triggered in a sharp increase in backup performance related questions. This was primarily around backups not completing in time or network bandwidth constraints in moving lots of data around. To make matters more difficult, each customer additionally had a different setup and required a detailed investigation into the backup configuration before assistance could be provided.
It made sense for Metasphere to design, build and recommend a standardised and fit-for-purpose backup process for their clients. Paul Harding, Senior Engineer at Metasphere, approached DSP– a leading Oracle support and development partner with a reputation for delivering innovative and cost effective solutions – for advice on how best to resolve the problem.
DSP worked with Paul to understand the application requirements and the nature of Metasphere’s client support requests, before implementing an incremental RMAN backup solution that could be rolled out across Metasphere’s client base.
Phase one of the solution switched from a 'daily full' back up to a 'weekly full' and 'daily incremental' backup. The solution was then formalised in clearly understood and commented scripts that are proven on multiple client sites, providing:
- Granularity of control
- Highest level of DR in the event of a major outage
- Improved performance across the network
DSP had experienced Metasphere's issues before and was therefore able to provide a solution that had already been proven in multiple production environments. As a result the team were able to free up key individuals to continue to focus on product development, rather than repeatedly devoting their efforts to supporting a poorly performing process.
With a solid automation process now in place, the backups are created significantly quicker and the resulting daily network transfers are much smaller in size. Backups are now transferred to an off-site location using the same network that had previously been saturated by the original backup process, eliminating the need for additional storage.