Uploaded image for project: 'dcm4chee'
  1. DCMEE-1881

Image received in dcm4chee causes 100% CPU usage when an attempt is made to view or retrieve it via wado


    • Type: Bug
    • Status: Open
    • Priority: Minor
    • Resolution: Unresolved
    • Affects Version/s: dcm4chee-2.17.1
    • Fix Version/s: None
    • Component/s: WADO
    • Labels:
    • Environment:
      dcm4chee 2.17.1, pgsql, CentOS 6.3
    • Tracking Status:
      Risk Analysis - Todo, Test Spec - ToReview, Test State - Not tested


      I received an image which was successfully received and stored in dcm4chee, as well as forwarded onto a different dcm4chee instance.

      However, when I try to view the image through the dcm4chee-web3 interface, or retrieve it via wado - it causes 100% CPU usage on the system. I believe it is because the wado requests hang, and lock up in some endless loop. This is visible in the jmx-console -> Tomcat Status, where one can see the increasing number of wado requests for that image (one each time a view or retrieve attempt is made). The CPU is fully used by the dcm4chee java process, until the system becomes unusable.

      The only solution is to stop and start the dcm4chee process, and to delete the image off the system.

      The image imports successfully into OsiriX, but the pixel data looks to be somewhat corrupt in some way. Even if the image is corrupt (which I do believe it is), it seems that no image should be able to bring down the whole PACS, but maybe some check needs to be put in the wado service which would handle this condition, or limit the amount of resources/time that a wado service thread can run when serving up a dicom object?...




            • Assignee:
              gunter Gunter Zeilinger
              jkoj Jordan K
            • Votes:
              0 Vote for this issue
              0 Start watching this issue


              • Created: