FHIR Chat · build.fhir.org broken

Stream: IG creation

Topic: build.fhir.org broken

Max Masnick (Jan 02 2022 at 16:09):

It looks like build.fhir.org is failing due to packages not loading: http://build.fhir.org/ig/HL7/fhir-shc-vaccination-ig/branches/master/failure/build.log

Sushi: error Failed to load us.nlm.vsac#0.4.0: The package us.nlm.vsac#0.4.0 could not be loaded locally or from the FHIR package registry. (00:01.0555)
Sushi: error Failed to load us.cdc.phinvads#0.7.0: The package us.cdc.phinvads#0.7.0 could not be loaded locally or from the FHIR package registry. (00:01.0570)
Sushi: error Failed to load hl7.fhir.r4.core#4.0.1: The package hl7.fhir.r4.core#4.0.1 could not be loaded locally or from the FHIR package registry. (00:01.0587)
Sushi: error Valid StructureDefinition resource not found. The FHIR package in your local cache may be corrupt. Local FHIR cache can be found at <home-directory>/.fhir/packages. For more information, see https://wiki.hl7.org/FHIR_Package_Cache#Location. (00:01.0674)

I cleared .fhir locally, and I cannot reproduce this SUSHI (i.e., the packages server is working).

Max Masnick (Jan 02 2022 at 20:30):

I'm now seeing this issue with our unit tests run via GitHub Actions (which also rely on SUSHI). So it looks like something may be amiss with wherever the packages are supposed to be downloaded from.

Kevin Power (Jan 03 2022 at 14:16):

Looks like there are still problems:

Sushi: error Failed to load hl7.fhir.r4.core#4.0.1: The package hl7.fhir.r4.core#4.0.1 could not be loaded locally or from the FHIR package registry. (00:01.0874)
Sushi: info  Found 1 file in an input/* resource folder that was neither XML nor JSON. This file was not processed as a resource by SUSHI. (00:02.0412)
Sushi: error Valid StructureDefinition resource not found. The FHIR package in your local cache may be corrupt. Local FHIR cache can be found at <home-directory>/.fhir/packages. For more information, see https://wiki.hl7.org/FHIR_Package_Cache#Location. (00:02.0414)

Kevin Power (Jan 03 2022 at 14:20):

http://build.fhir.org/ig/HL7/genomics-reporting/branches/master/failure/build.log

Oliver Egger (Jan 03 2022 at 14:24):

i think this might be an issue with sushi, see https://chat.fhir.org/#narrow/stream/215610-shorthand/topic/R5.20support

Chris Moesel (Jan 03 2022 at 14:40):

@Grahame Grieve -- did you (or someone) make changes to the FHIR package registries? Are packages not served at the same locations they used to be served? I haven't had time to investigate yet (meetings right now), but trying to see if anyone is aware of changes to those servers that might be the cause.

Chris Moesel (Jan 03 2022 at 14:42):

Or @Ward Weistra -- do you know?

Vassil Peytchev (Jan 03 2022 at 15:00):

build.fhir.org has an expired certificate - would that cause a problem?

Ward Weistra (Jan 03 2022 at 15:00):

Indeed, that's also what I'm seeing since just now: https://packages.fhir.org/catalog?name=hl7.fhir.us.core

Jens Villadsen (Jan 03 2022 at 15:06):

image.png

Jens Villadsen (Jan 03 2022 at 15:06):

oh yes ...

Chris Moesel (Jan 03 2022 at 15:24):

Ah! That would do it! SUSHI uses HTTPS and will fail if the certificate is invalid (as in this case). Who do we need to talk to in order to get that certificate updated?

Lynn Laakso (Jan 03 2022 at 15:30):

Checked with HQ and it sounds like it's Grahame

Melva Peters (Jan 03 2022 at 15:54):

@Grahame Grieve

Chris Moesel (Jan 03 2022 at 18:51):

For anyone who is absolutely stuck and needs to get local builds working (for SUSHI), I've posted a temporary workaround on the shorthand stream here. Please note the included caveats!