-1

Set up Nvidia Clara on a AWS instance in through Kubernetes.

I run my dicom adapter through clara and it is running on POD. I activate the listener by running storescp -v --fork -aet DCM4CHEE 104 which shows as listening

If i locally run storescu localhost 104 <image.dcm> it works fine.

But if I try to access the POD (type loadbalancer) using its external facing IP address

./storescu <external_IP_of_pod> <port> <image.dcm>

I receive the following error

D: ====================== BEGIN A-ASSOCIATE-RQ =====================
D: Our Implementation Class UID:      1.2.276.0.7230010.3.0.3.6.5
D: Our Implementation Version Name:   OFFIS_DCMTK_365
D: Their Implementation Class UID:
D: Their Implementation Version Name:
D: Application Context Name:    1.2.840.10008.3.1.1.1
D: Calling Application Name:    STORESCU
D: Called Application Name:     ANY-SCP
D: Responding Application Name: ANY-SCP
D: Our Max PDU Receive Size:    16384
D: Their Max PDU Receive Size:  0
D: Presentation Contexts:
D:   Context ID:        1 (Proposed)
D:     Abstract Syntax: =AmbulatoryECGWaveformStorage
D:     Proposed SCP/SCU Role: Default
D:     Proposed Transfer Syntax(es):
D:       =LittleEndianExplicit

......

Requested Extended Negotiation: none
D: Accepted Extended Negotiation:  none
D: Requested User Identity Negotiation: none
D: User Identity Negotiation Response:  none
D: ======================= END A-ASSOCIATE-RQ ======================
I: Requesting Association
D: setting network send timeout to 60 seconds
D: setting network receive timeout to 60 seconds
D: Constructing Associate RQ PDU
F: Association Rejected:
F: Result: Rejected Permanent, Source: Service User
F: Reason: Called AE Title Not Recognized

Been stuck for a while, any help will be greatly appreciated! thanks

The BrownBatman
  • 2,593
  • 1
  • 13
  • 29

2 Answers2

2

Following is the line from the log:

Called Application Name: ANY-SCP

This indicates that the Called AE Title set on SCU is ANY-SCP.

Now, look at the last line of the log:

Called AE Title Not Recognized

That means the Called AE Title you set on SCU (ANY-SCP) is not matching with what is set on SCP.

I do not know the toolkit but I guess SCP AE Title is set to DCM4CHEE looking at your command code. You have to configure your SCU to set the same as Called AE Title. This should resolve the issue.

If the same configuration works on same system but fails when connecting from different system and load balancer is introduced, may be that your request is being sent/forwarded/redirected to other SCP that you do not expect. The AE Title does not match there (as expected) and hence the error. In this case, it is more a network issue than a DICOM issue.

Alternatively, I am not sure but if you can somehow configure the SCP to ignore the AE Title mismatch???? Anyway, I will not recommend this.

Amit Joshi
  • 15,448
  • 21
  • 77
  • 141
  • Thanks for your input - will definitely try this out and get back to you! – The BrownBatman Nov 14 '19 at 09:30
  • not really, I got the example clara pipeline running no problem but then i create my own pipeline it gives me the same error. Even though now its calling the same Application Name i type in... – The BrownBatman Nov 19 '19 at 21:20
0

You need to know the AET of the SCP you are trying to connect to.

Also, the SCP may have a configuration table that specify which SCU AET can connect to it.

./storescu [-aet <SCU AET>] -aec <SCP AET> <external_IP_of_pod> <port> <image.dcm>
Paolo Brandoli
  • 4,681
  • 26
  • 38