Posts by marcelvanherk

If its the patientid the behavior (reject) is right, but the errormessage is incorrect

Hi, thanks for testing.

Can you tell me what they changed? In your log you see that conquest reports "removefiles 0 images" i.e., it tries to remove the existing image and fails.

Marcel

Hi,

this must be a bug, it is not supposed to try and write the same data twice. Can you redo this with debug log on for me and post the full log?

Marcel

Thanks and good idea, but not a very high priority.

Marcel

Great!

Marcel

Never got around to deciding it.

No,

it is hardcoded to retry 10 times and wait a second inbetween.

cheers

Marcel

Hi, just been testing. This webpage works:

<FORM ACTION="/api/dicom/rs/attach" METHOD=POST ENCTYPE="multipart/form-data">
<INPUT NAME=script  TYPE=HIDDEN VALUE=reject>
Upload file to enter into server (dcm/v2/HL7/zip/7z/gz/tar): <INPUT NAME=filetoupload SIZE=40 TYPE=file VALUE=>
<INPUT TYPE=SUBMIT VALUE=Go>
</FORM>
<FORM ACTION="/api/dicom/rs/attachdicom" METHOD=POST ENCTYPE="multipart/form-data">
<INPUT NAME=script  TYPE=HIDDEN VALUE=Data:AddImage();return(1)>
Upload dicom file to enter into server (dcm): <INPUT NAME=filetoupload SIZE=40 TYPE=file VALUE=>
<INPUT TYPE=SUBMIT VALUE=Go>
</FORM>

So, I guess there is something wrong with your upload code. You can have a look in a browser to see what my example sends.

Put the tests into test.html on github, and catch empty servercommand() in posters.lua

Marcel

And I put the retrying version on github ;->>>

Marcel

I meant option 2!

A stopgap solution is to script a delay e.g. by defining dgate64.lua (if your binary is dgate64.exe) as follows:

print('delay')
t=os.time()
while(os.time()-t<5) do end
print('starting')

This will wait 5s before the monitor thread is activated.

Marcel

Hm,

it would not be an easy change to do this at all. The disk storage is for failed forward attempts. You could use lua code to keep track of all studies that are received or use AccessTime in the database to know which studies have just been added so that these can be retried. I.e. by sending these to itself.

Marcel

Hi,

The incoming folder thread (function monitorthread) opens a database. I guess that can retry e.g. 10s.

Marcel

Hi, I will have a look. There is disk code, but I forgot how to activate it.

Super!

Marcel

And can you check the memory usage of the server and database?

Hm,

Not seen this before. Can you also post the postgres errors?

Marcel

Is there an error message ?

Thanks a lot!

Does this fix both? If you approve I will merge into github

Marcel

Ok,

the sql database automatically stores any non-sequence item that is listed in dicom.sql. You can maybe sort your problem by using lua to create private elements to the RTPLAN object and define these in dicom.sql so that they are written. However, as I pointed out there can be hundreds of control points so I am not sure how you would code that in a single text string. I think database fields in conquest are limited to 256 characters.

A better solution is to create a database table yourself (you can use lua to do this) and use lua sql() command to write into that and dbquery() to read.

console.bat runs a simple console where you can try any lua code. It stores the history of what you typed.

A bit of example code (from overview.lua in the distribution, it has examples for all lua functions):

sql("CREATE TABLE UserTable (CounterID varchar(99), Val integer)");

sql("INSERT INTO UserTable (CounterID,Val) VALUES ('CT',1) ON DUPLICATE KEY UPDATE Val=Val+1");

Marcel

Good,

You show that you read any information from a plan. Can you now explain what you actually want to achieve?

Marcel