Stream: IG creation
Topic: Publisher 1.1.0 build is hanging
Rob Hausam (Jul 01 2020 at 17:06):
Trying to build the IPS IG with Publisher 1.1.0, the build is hanging without any errors or warnings:
RHausamMacbookPro:fhir-ips rhausam$ ./_genonce.sh
Checking internet connection...
Online
FHIR IG Publisher Version 1.1.0 (Git# e654d867a231). Built 2020-07-01T02:23:14.763Z (12 hours old)
File log: /var/folders/vv/4dq6q1157n3b62yb7cxyr6br0000gn/T/fhir-ig-publisher-tmp.log
Detected Java version: 14.0.1 from /Library/Java/JavaVirtualMachines/adoptopenjdk-14.jdk/Contents/Home on x86_64 (64bit). 4096MB available
Parameters: -ig ig.ini
dir = /Users/rhausam/git-repo/fhir-ips, path = /usr/local/Cellar/pyenv-virtualenv/1.1.5/shims:/Users/rhausam/.pyenv/shims:/Library/Frameworks/Python.framework/Versions/2.7/bin:/usr/local/opt/bison/bin:/usr/local/opt/ruby/bin:/opt/local/bin:/opt/local/sbin:/Users/rhausam/google-cloud-sdk/bin:/usr/local/bin:/usr/bin:/bin:/usr/sbin:/sbin:/Applications/VMware Fusion.app/Contents/Public:/usr/local/MacGPG2/bin:/usr/local/share/dotnet:/opt/X11/bin:~/.dotnet/tools:/usr/local/git/bin:/opt/apache-maven-3.6.2/bin:/usr/local/ant/bin
Run time = Wednesday, July 1, 2020 at 9:59:53 AM Central Daylight Time (2020-07-01T09:59:53-05:00)
Package Cache: /Users/rhausam/.fhir/packages (00:00.0072)
Load Template from hl7.fhir.template (00:01.0028)
Load Template from hl7.base.template#current (00:03.0717)
Load Template from fhir.base.template#current (00:03.0821)
onLoad.findSpreadsheets:
onLoad.updateIg:
[xslt] Processing /Users/rhausam/git-repo/fhir-ips/template/onLoad-ig-working.xml to /Users/rhausam/git-repo/fhir-ips/template/onLoad-ig-updated.xml
[xslt] Loading stylesheet /Users/rhausam/git-repo/fhir-ips/template/scripts/onLoad.xslt
onLoad.extend:
onLoad:
Root directory: /Users/rhausam/git-repo/fhir-ips (00:07.0920)
Core Package hl7.fhir.r4.core#4.0.1
Terminology Cache is at /Users/rhausam/git-repo/fhir-ips/input-cache/txcache. 7 files in cache (00:17.0203)
Connect to Terminology Server at http://tx.fhir.org (00:20.0236)
Load Package hl7.fhir.pubpack#0.0.6
Load Package hl7.fhir.xver-extensions#0.0.4
Initialization complete (00:22.0506)
Load Content (00:22.0507)
Translating CQL source (00:22.0596)
Translating CQL source in folder /Users/rhausam/git-repo/fhir-ips/input/cql (00:22.0597)
Processing Conformance Resources (00:29.0747)
Terminology server: Check for supported code systems for http://unstats.un.org/unsd/methods/m49/m49.htm (00:33.0693)
It appears to hang at the terminology server validation. But the tx.fhir.org server itself is up, and restarting it didn't make a difference.
Rob Hausam (Jul 01 2020 at 17:16):
The tx server is able to expand the http://hl7.org/fhir/ValueSet/jurisdiction value set (where the http://unstats.un.org/unsd/methods/m49/m49.htm code system is used) without a problem. So I'm not seeing an obvious reason so far why the publisher isn't getting any further with the build.
Grahame Grieve (Jul 01 2020 at 20:09):
it won't be that - that's just the last thing it did.
Grahame Grieve (Jul 01 2020 at 20:10):
How long did you give it?
Rob Hausam (Jul 01 2020 at 20:10):
at least 30 minutes
Rob Hausam (Jul 01 2020 at 20:12):
the next step is to look to see what it was attempting to do next
Rob Hausam (Jul 01 2020 at 20:16):
Actually, I just looked at my last attempt - and it turns out that it did complete. :) I'll check to see how long that step really took (it was a long time).
Rob Hausam (Jul 01 2020 at 20:27):
Well, this is really interesting! - sort of :( It turns out that from the last step that I showed above, to the next step of Generating Narratives - it took a total of 3 hours, 6 minutes and 7 seconds. :( @Grahame Grieve?
Grahame Grieve (Jul 01 2020 at 20:27):
it probably won't take that long next time, and if it doesn't, a diff on the txcache will tell you what happened
Rob Hausam (Jul 01 2020 at 20:35):
OK, I'm sure you're right about that, regarding subsequent builds. I'm pretty sure that I'm as aware as probably anyone that building the txcache can take a long time. But, as I recall, it normally gives output for each terminology artifact that it processes. Without that and with no response at all for over 3 hours, how would anyone expect that it was doing its normal job and would actually complete (or have any idea when that would happen)?
Grahame Grieve (Jul 01 2020 at 20:40):
I'm not sure. Does txCache tell you anything?
Rob Hausam (Jul 01 2020 at 20:46):
I'll look. The second build was 5 min and 39 sec. total.
Rob Hausam (Jul 01 2020 at 21:03):
The txcache looks pretty unremarkable to me. The snomed cache is 20.2 MB, which is pretty big - but I don't recall if that's any bigger than usual, or why that would be a problem.
Grahame Grieve (Jul 01 2020 at 21:20):
you can do a diff?
Last updated: Apr 12 2022 at 19:14 UTC
