Binary requests time out

86 views
Skip to first unread message

Donald French

unread,
Jun 24, 2019, 3:45:44 PM6/24/19
to Cerner FHIR Developers
  • Issue Summary: When requesting a Binary for a DocumentReference they "all" return 504 Endpoint request timed out. This was working last week at the workshop.  I have had this other random times.  
  • X-Request-Id or CorrelationId: x-amzn-RequestId →0f0e12b9-96b8-11e9-9eec-db219749b9de
Thank you

Donald French

unread,
Jun 27, 2019, 10:03:20 PM6/27/19
to Cerner FHIR Developers
THis is still an issue. Is anybody else having the same problem? The problem is easily repeated with Nancy's documents.  I can get Goldberg's.  Is it timing out because the generation of the 60-page documents takes too long?  If so this is a critical problem.

Benjamin Eichhorn (Cerner)

unread,
Jun 28, 2019, 12:50:35 PM6/28/19
to cerner-fhir...@googlegroups.com
Hi Donald,

Apologies for the delay. It does appear that the request you made took longer than our timeout period which is why you received the 504. Depending on the size of the document it can take a longer time to process and generate. You can attempt to retry the request tor receive the document in question. 

Thanks,
Ben (Cerner)

Donald French

unread,
Jun 28, 2019, 1:08:15 PM6/28/19
to cerner-fhir...@googlegroups.com
I have retried several times on documents several times. I suspected that was the issue and asked at the class and was told the same thing. 

The problem is two fold. One, the samples for Nancy and many others are over 50 pages long. I can find none in the 1200+ documents that are reasonable size and usable 

Two it shows that you do not support large documents. 

Thank you
Don French
 

From: 'Benjamin Eichhorn (Cerner)' via Cerner FHIR Developers <cerner-fhir...@googlegroups.com>
Sent: Friday, June 28, 2019 9:50 AM
To: Cerner FHIR Developers
Subject: Re: Binary requests time out
 
Hi Donald,

Apologies for the delay. It does appear that the request you made took longer than our timeout period which is why you received the 503. Depending on the size of the document it can take a longer time to process and generate. You can attempt to retry the request tor receive the document in question. 

Thanks,
Ben (Cerner)

--
You received this message because you are subscribed to a topic in the Google Groups "Cerner FHIR Developers" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/cerner-fhir-developers/EUU9RBvduDI/unsubscribe.
To unsubscribe from this group and all its topics, send an email to cerner-fhir-devel...@googlegroups.com.
To post to this group, send email to cerner-fhir...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cerner-fhir-developers/29ae1ff3-9e8d-4a66-af31-7ac49b0bd15c%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Benjamin Eichhorn (Cerner)

unread,
Jun 28, 2019, 1:19:59 PM6/28/19
to Cerner FHIR Developers
Hi Donald,

I understand the issue. Please understand that at this time we are aware that it can be difficult to retrieve larger documents and while this doesn't help immediately, we do have work to enable documents to be retrieved faster and more consistently. 

I sincerely apologize for the inconvenience,
Ben (Cerner)
 

Donald French

unread,
Jun 28, 2019, 1:44:58 PM6/28/19
to cerner-fhir...@googlegroups.com
Thank you. Can you put out some smaller documents that can be used?

Don French

PS: I do appreciate your work.
--
You received this message because you are subscribed to a topic in the Google Groups "Cerner FHIR Developers" group.
To unsubscribe from this group and all its topics, send an email to cerner-fhir-devel...@googlegroups.com.
To post to this group, send email to cerner-fhir...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
Sent from Mailspring

Benjamin Eichhorn (Cerner)

unread,
Jun 28, 2019, 5:05:59 PM6/28/19
to Cerner FHIR Developers
Hi Donald,

I've had several documents added onto "Jason Mendoza" in FHIRPLAY (see link below). He should have a few examples of common notes we would see on DocumentReference.


Thanks,
Ben (Cerner)

Donald French

unread,
Jun 28, 2019, 5:13:23 PM6/28/19
to cerner-fhir...@googlegroups.com
Thank you. I was able to access the document I tested.  However 10 seconds for a 2 page document seems excessive. Is that the norm for your fhir interface?

Are scanned documents DocumentReferecences or DiagnosticReports?


Don French
--
You received this message because you are subscribed to a topic in the Google Groups "Cerner FHIR Developers" group.
To unsubscribe from this group and all its topics, send an email to cerner-fhir-devel...@googlegroups.com.
To post to this group, send email to cerner-fhir...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
Sent from Mailspring

Joshua Fisk (Cerner)

unread,
Jul 1, 2019, 2:58:55 PM7/1/19
to Cerner FHIR Developers
The time it takes to retrieve documents is going to depend mostly on the hardware at the client site.
 
As for scanned documents the short answer is that it depends. If the scanned document is configured to be saved as a clinical note it would only be retrievable through the DocumentReference endpoint. Our DiagnosticReport endpoint only returns documents which have been mapped to the DiagnosticReport.category of http://hl7.org/fhir/v2/0074|RAD. Also keep in mind that any documents retrieved as a PDF will have a client defined header and footer applied to them.

Joshua Fisk

Donald French

unread,
Jul 1, 2019, 6:31:35 PM7/1/19
to cerner-fhir...@googlegroups.com
This is a very high powered server accessing the Document references. I doubt that is the cause other apis I talk to have no problem with larger documents.  

Don French

--
You received this message because you are subscribed to a topic in the Google Groups "Cerner FHIR Developers" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/cerner-fhir-developers/EUU9RBvduDI/unsubscribe.
To unsubscribe from this group and all its topics, send an email to cerner-fhir-devel...@googlegroups.com.
To post to this group, send email to cerner-fhir...@googlegroups.com.

Joshua Fisk (Cerner)

unread,
Jul 2, 2019, 11:40:37 AM7/2/19
to Cerner FHIR Developers
Hi Don,

The client site hardware that I referred to is the Cerner hardware for that site. If you are using Cerner's sandbox then the hardware in question is the Cerner hardware that backs the free public sandbox and as such it is not production tier quality. In actual client site implementations the hardware and configuration would have much better load capabilities and availability.

- Joshua Fisk

On Monday, July 1, 2019 at 5:31:35 PM UTC-5, Donald French wrote:
This is a very high powered server accessing the Document references. I doubt that is the cause other apis I talk to have no problem with larger documents.  

Don French

On Mon, Jul 1, 2019 at 11:58 AM, Joshua Fisk (Cerner) <joshu...@gmail.com> wrote:
The time it takes to retrieve documents is going to depend mostly on the hardware at the client site.
As for scanned documents the short answer is that it depends. If the scanned document is configured to be saved as a clinical note it would only be retrievable through the DocumentReference endpoint. Our DiagnosticReport endpoint only returns documents which have been mapped to the DiagnosticReport.category of http://hl7.org/fhir/v2/0074|RAD. Also keep in mind that any documents retrieved as a PDF will have a client defined header and footer applied to them.

Joshua Fisk

--
You received this message because you are subscribed to a topic in the Google Groups "Cerner FHIR Developers" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/cerner-fhir-developers/EUU9RBvduDI/unsubscribe.
To unsubscribe from this group and all its topics, send an email to cerner-fhir-developers+unsub...@googlegroups.com.

Donald French

unread,
Jul 2, 2019, 11:46:22 AM7/2/19
to cerner-fhir...@googlegroups.com, Cerner FHIR Developers
Thank you for the clarification. I assumed the sandbox hardware was small, old, and slow. 

I discussed it with one of my clients and the say because they are on the Citrix/cloud version their XR requests are slow and time out so I was not sure. 

To unsubscribe from this group and all its topics, send an email to cerner-fhir-devel...@googlegroups.com.

To post to this group, send email to cerner-fhir...@googlegroups.com.
Reply all
Reply to author
Forward
0 new messages