Clarification regarding the imagectn configuration file imagectn.cfg

All other questions regarding DCMTK

Moderator: Moderator Team

Post Reply
Message
Author
Jyoti
Posts: 17
Joined: Wed, 2004-12-01, 04:33

Clarification regarding the imagectn configuration file imagectn.cfg

#1 Post by Jyoti » Wed, 2004-12-01, 04:55

Hi,

I am new to DICOM and DCMTK and have just installed DCMTk and started using it.

I am a bit confused regarding the parameter HostTable parameter of the file imagectn.cfg. According to my understanding this table lists the hosts which are allowed to access my Image database. Now I am not clear on why do we need to have the ( AETitle, HostName, Portnumber ) triplet for each host? Does this say that I can access only the storage area symbolised by the name AETitle on the host Hostname which has its imagectn running on port Portnumber ? Please correct me if I am wrong.

thanks,
Jyoti

Marco Eichelberg
OFFIS DICOM Team
OFFIS DICOM Team
Posts: 1280
Joined: Tue, 2004-11-02, 17:22
Location: Oldenburg, Germany
Contact:

#2 Post by Marco Eichelberg » Wed, 2004-12-01, 10:16

The DICOM Query/Retrieve protocol uses two separate connections for image downloads from the archive. The first connection is initiated by the client and is used to transmit the query and retrieve requests and status update messages. The second connection is initiated from the archive to the client (!) and is used to deliver the images.
This means that the archive needs to know the host name (or IP address) and port number under which the client will accept the second connection. The client itself only transmits a symbolic AETITLE, and the table in imagectn.cfg is used to look up the hostname and port number from that.

Jyoti
Posts: 17
Joined: Wed, 2004-12-01, 04:33

#3 Post by Jyoti » Wed, 2004-12-01, 11:28

Marco Eichelberg wrote:The DICOM Query/Retrieve protocol uses two separate connections for image downloads from the archive. The first connection is initiated by the client and is used to transmit the query and retrieve requests and status update messages. The second connection is initiated from the archive to the client (!) and is used to deliver the images.
This means that the archive needs to know the host name (or IP address) and port number under which the client will accept the second connection. The client itself only transmits a symbolic AETITLE, and the table in imagectn.cfg is used to look up the hostname and port number from that.
Your reply has clarified most of the things for me. So what I understand from this is that the imagectn service makes use of the imagectn.cfg to know the port and hostname of the client from which a request comes. This is for starting another connection to the client to transfer images as requested by the client.

A few questions here:

1. Does the connection from the server to the client start automatically as soon as a request for images arrives from the client?

2. Which AETitle does the client transfer? The one which symbolises the client archive or the one which symbolises the server archive? If it is the one that symbolises the server archive, then how will a server to client connection start if the access for it is set to ANY? In that case we can't map ANY to a hostname from the HostTable to get the port details.

- Jyoti

Marco Eichelberg
OFFIS DICOM Team
OFFIS DICOM Team
Posts: 1280
Joined: Tue, 2004-11-02, 17:22
Location: Oldenburg, Germany
Contact:

#4 Post by Marco Eichelberg » Wed, 2004-12-01, 11:37

The connection from the server to the clients starts as soon as a well-formed retrieve request (C-MOVE-RQ message) is received. Queries against the index database do not require the additional connection.
In the retrieve request, the client transmits a third AETitle, called the "Move AE Title". This is resolved into hostname and port number. This means that DICOM Query/Retrieve is essentially a three-point protocol. There is no need that the Move AETitle is identical to the client initiating the retrieval, i. e. the client can request the archive to send images "somewhere else".

Jyoti
Posts: 17
Joined: Wed, 2004-12-01, 04:33

#5 Post by Jyoti » Thu, 2004-12-02, 12:11

I just have a few more question:

1. Does the client make use of the imagectn.cfg file present in the client's database or is it only for the server to validate the clients? If yes, for what?

2. Is there a way for the client to find out what all repository servers it can access and at which port these servers have their imagectn application running?

Thanks for your help,
Jyoti

Jörg Riesmeier
ICSMED DICOM Services
ICSMED DICOM Services
Posts: 2217
Joined: Fri, 2004-10-29, 21:38
Location: Oldenburg, Germany

#6 Post by Jörg Riesmeier » Thu, 2004-12-02, 12:55

The answer to question 1 is: No, the Q/R client does not use the "imagectn.cfg". For details see "ctnconf.txt".

The answer to question 2 is: No, there is no way for the client to determine this information automatically. Maybe, the new DICOM configuration management service is what you're looking for (see DICOM 2004 standard for details).

Post Reply

Who is online

Users browsing this forum: Google [Bot] and 1 guest