auto forward

  • Marcel,


    I am autoforwarding from two different Conquest servers. I put in as you suggested the line:


    forward series to TIAPACS


    since both are very busy servers and have a lot of large series. However, it takes a very long time to forward the series. I amtalking up to 24 hours. is there a way to shorten the time frame it waits before auto forwarding?


    thanks.


    Pat Stacey

  • maybe it is the amount of studies??? we do have a lot on both servers. one is a major hospital and the other is the biggest imaging center in Western NY. 10 minutes is too long to wait in the queue. Especially for stat reads. Is there a way to change that time to maybe at the most 5 min delay, maybe even shorter??


    Thanks.

  • Sorry I did not get back to you sooner, just too busy.


    However, my forward delay was at 300. I still had the delay that was much longer so I changed it to 120. Seems to have helped, but still taking a long time to forward. How big is the queue? Is there a size limit? Most of the studies are very large CT and MRI and lots and lots of US and CR


    Pat

  • Hi Pat,


    the queue size can be changed as well. For "Forward to" the queue has one entry per slices. For "forward series to" there is a separate queue with one entry per series. The latter queue is also used for prefetch, preretrieve, and other delayed forwarders (forward study, patient). I would not expect to run out of queue space for "forward series", but that may be your problem. If the queue is full the server blocks until there is space.


    QueueSize. This is the size (in entries) of the in-memory queues for mirror copies and exportconverters, default 128. Each entry takes 1.5k (per export converter) or 2k (for the mirror copy queue). Conquest addition.


    Is there anything the log would show me? Maybe you can mail me a zipped log of a day or so.


    Marcel

  • It also comes up with "starting server keep on trying delayed fetch/forward failures in DelayedFetchForwardFailures104?" when I start the gui. I tried to attach the log, but it is 4 megs, beyond the limit. Here it is 9:30 pm and studies from 10:30 am are just now getting out. Now that the data is no longer coming in, it is sending everything. This system is a Core 2 Intel, Windows XP Pro, 2 gigs ram, dual 500 mirrored hard drives. Our broadband is 10 down, 2 up and is the best we can get at this location.


    Something seems to bottle neck it up and I don't know why. The other Conquest is faster at getting the data over and it is at a major hospital with larger and much more files constantly coming in. That one is also a Core 2 Intel, 2 Gigs ram, same hard drives, Windows Server 2003. But, the broadband there is 20 up and down (fiber). Do you think that is the issue?


    Would opening up the queue work?


    Thanks.

  • I would like to interject a question into this thread too....it seems an appropriate thread.


    I have some times when the auto-forwarding queue takes a long time until the study will send. I assume it sends
    just one exam at a time, which is fine with me, as the receiving system is slow. But, even though I have
    a 5 minute collect delay, there are times that so many exams are queued up that 30 or 45 minutes later it may not have yet started to send a particular exam.


    Is there a way for me to find out what studies are queued up and waiting, so I can see how long the list is?


    Tim

  • Marcel, I seem to have tracked down the time. I have a 2 minute forward delay set. However, it is taking 20 minutes or longer to start the send. Now, we do have a lot of studies and they are coming in fast and furious. It is not unusual to see studies come in every 30 seconds or so and they are large. Is there a way to see the queue? increase it?


    Thanks.

  • Hi Pat,


    did you try to set like:


    QueueSize = 1000


    in dicom.ini?


    Soory, there is no viewing options for the queues (yet). And only failures get written to disk, so you cannot inspect it there either.


    Marcel

Participate now!

Don’t have an account yet? Register yourself now and be a part of our community!