The data server for the observation data have been moved to a `twin' operations database server. The objective is to keep an exact duplicate of that critical machine, to minimize the possibility of data loss and observing time lost due to hardware failures. Both servers work in a ``virtual disk'' mode, meaning that the database is saved on both machines simultaneously. The second objective (achieving HA - High Availability) was tested with satisfactory results for most circumstances. Some work is needed to make this functional for all cases and then we can ensure continued service, even in the case of partial or total hardware failure of the primary server.
Over the past few years we've been gradually renewing the hardware on our desktops & and servers. It the moment we just have received replacement for our main ``home account & mail storage'' server. It's also planned to decommission and replace a number of older desktop machines (for administrative & visitor use) with more modern ones.
Also, a number of aging desktop machines haven't been replaced until now because they are used for operation-critical missions (post-processing, etc), meaning that they're in a ``frozen'' status as we can't afford to just upgrade them without risking incompatibility issues. The advance on the new post-processing system (and other systems) will allow us to also upgrade/replace those computers, providing a more modern environment for the whole NOT.
Thomas Augusteijn 2010-02-09