I am working on the creation of an NHS / UK specific paper on ‘Image Availability’ which was the core message we promoted at RSNA last week. (Original paper is here http://tinyurl.com/a7xzfwb). For part of this, I would like to hear if any of you have concerns, suggestions or anecdotes about how the UK is using tools like a VNA to make sure images are available in the long term.
In a recent example, GE are warning us that replication isn’t protecting their PACS environments enough (see article on HealthData Management: http://goo.gl/TdlHt). We believe this applies to most PACS or related system, not just GE (although, in many respects, GE should be applauded for making us aware of the issue). Is this something that you have been made aware of? Is this situation something that you expect IT to take care of?
I welcome any comments, feedback or suggestions you might have on this or any other scenarios concerning ‘Image Availability’.
Hi Jamie.. I think its a little unfair to suggest the GE example illustrates that "replication isn't protecting PACS enough". True its a GE specific issue, that has been duely notified. That doesn't mean that replication per se is bad - in fact its a concept that works well in many mission critical environments (it doesn't have to be rocket science either)
The white paper linked to in the comments on HealthDataManagement dips gently into concepts like deduplication, version control and hierarchical storage which are all valid to some extent (all can also be debated:
In terms of "image availability" - I don't see why replication + backups + *good operating processes* aren't sufficient - at least within a single implementastion The other elements to the archiving technologies are only really about the implementation & economics of the process, no?
Not that they aren't important too.
Personally, I'd be more concerned about making the implementation itself redundant - i.e. full stack redundancy - ensuring that flaws in hardware, OpSys, database, application, DICOM implementation etc are covered in the face of potential disruption
Martin, I agree with you that Replication + backup organised in a best practice way, will be sufficient, the issue that we see is that for whatever reason, this isn’t being achieved OR isn’t working and I would also agree that replication for high availability is a must.
What we are seeing daily with our customers and prospects is that this best practice isn’t working and even looking at the example quoted (and this isn’t something specific to that vendor) they use the words ‘data loss’ and I am sure that they implemented best practice.
Looking wider, and looking at the quotes / anecdotes I am hoping to find, what about recovery times? Again, most customers and prospects I talk to have concerns about the time it’s going to take to recover a system; if a recovery takes 6 weeks then is this adequate protection?