This site has a high availability setup with two redundant syngo.Plaza servers, combined in a so called server cluster.For the dicom-partners in the clinical network this cluster will appear and will be treated transparently as a single syngo.Plaza server.The intention of this setup is to ensure an operation of this single out-side visible syngo.Plaza server without interruption.To achieve this the "cluster manager" software (provided by microsoft) will monitor the state of the activate "operative" server for proper operation.In case this "operative" server should fail, the cluster manager will activate the second - "spare" server to take over.At any time only one of the servers exclusively will take care for the tasks to be done by the server cluster.At this site the cluster manager got into a disordered state caused by hardware problems.In this error state he frequently and rapidly switched between the two servers of the cluster.As a result the two servers at some points in time sporadically were running in parallel - not exclusively.During these timepoints the tasks to be performed by the server-cluster were done in parallel by both of the servers, writing concurrently onto the same data files and same database records.Depending on the non-predictable sequence of the actions in this error scenario, the content of data files and database references had mixed patient information.The syngo.Plaza is working as designed and intended.With this information, a death or serious injury is highly improbable.(b)(6).
|
Siemens became aware of an issue with the potential of data patient mix-up from a site in (b)(6).Per the complaint report, there was no injury associated with this issue.However, siemens factory experts evaluated the reported issue as low risk for the product and risk for the site.It was reported that patient data got mixed up for about 3 days during 2 occurrences.The mix-up was corrected and no impact for patients was reported.Some of the data mix-up was detected by the user.The complete amount of data mix-up was determined and all data was corrected.The syngo.Plaza product is working as designed and intended; the established mitigation to roll back a failed storing activity is currently in place and effective under normal circumstances.However, the implementation did not consider such an exceptional scenario as in this case.Therefore an investigation has started to determine whether the mitigation can be enhanced.
|