- Expose
/ga4gh/drs/v1/objects/{object_id}/access/{access_id}
GA4GH DRS Access API via Fence (#957) - Use RAS v1.1 Passport instead of RAS v1.0 passport (#956)
- Validate RAS v1.1 Passports (#956)
- Added logic to create Gen3 Passports (#956)
- Add registration endpoints (registration form endpoint and admin-only list endpoint) (#906)
- Use Azure Blob Storage with Fence (#946)
- A change in the gen3cirrus package for the get_signed_url function was made, the function parameters were not updated in fence and returned an invalid parameter error. Adding missing parameters to the function in Fence resolved error. (#955)
- Bug fix in checking email claims for Azure AD (in case the
email
claim is missing from the AAD response) (#946)
- remove deprecated project access from cookies (making them smaller) (#958)
- update to latest python base image (#959)
- Improvements around CSRF handling (#906)
- Test suite update (#946)
- New dependency Flask-WTF (#906)
- Python (#946)
azure-storage-blob=^12.6.0
(#946)azure-core==1.13.0
(#946)cryptography==2.8
(#946)isodate==0.6.0
(#946)msrest==0.6.21
(#946)oauth==3.1.0
(#946)requests-oauthlib==1.3.0
(#946)six==1.16.0
(#946)
- Requires these cloud-automation updates to revproxy: uc-cdis/cloud-automation#1580 (#906)
- Add
indexd
settings files for PostgreSQL SSL support (#317)
- Clean up test settings (#317)
- update to latest python base image (#322)
- Add notes for local development (#315)
- Add example Azure DevOps pipeline and Documentation (#316)
- Include example secrets to load PostgreSQL settings for
indexd
(#317)
- adding aligned_reads_index to the extra nodes on VCF file exports to make them inline with CRAM files (#58)
- Fix encoding issue when exporting (#57)
- Handle exceptions without 'message' attribute (#178)
- update to latest python base image (#179)
- Do not crash when Peregrine returns a non-200 code, and try to parse the data anyway (#40)
- add vaccine map layer (#908)
- Add contentSecurityPolicyConnectSRCWhiteList to portal config to allow for urls to be whitelisted by Content-Security-Policy (#919)
- Explorer: Redirects users to the login page if they try to download the data without logging in. (#912)
- Closable banner at the top of workspace page to remind users to link account to other data commons (#916)
- add an R logo for workspaces with titles formatted " *R Notebook *" (#915)
- (#915)
- NCT: Add a web form component to allow signing up for email lists at govdelivery.com (#911)
- A user can select one or more studies to request for download. This dispatches a sower job which is monitored until it either sends back an s3 presigned url, or fails with an error description. (#906)
- Mitigated PFB export bug (PXP-8400) by increasing timeout before checking for status of PFB export job (#871)
- Fix occ data bucket from being blocked by Content Security Policy (#919)
- Workspace: fixed a bug causing workspace be putting into incorrect
Launching
state if some containers has already been terminated (#903)
- refactor map data handling logic (#908)
- update colors of Mobility Data layers (#908)
- Misc: add the optional
Discovery
page tonct
andcovid19
bundles (note theworkspace
bundle still don't have it) (#918) - New configurable subtitle helper text under the Discovery page search box in the gitops.json attribute inputSubtitle. (#909)
- UX improvements in the Discovery page's Data Availability column header -- 3 filter options, remove sort icon (#909)
- Removing
AccessLevel.UNACCESSIBLE
from table sorting and filtering options (#909) - Core metadata / download page: remove empty values from the header (#910)
- Misc: add option for bundle name:
ecosystem
, which is the same asheal
bundle under the hood (#918) - Requires
loginForDownload
flag set to true in gitops (#912) - User registration before download feature depends on uc-cdis/fence#906 (#912)
- Requires the configuration of the
batch-export
sower job: uc-cdis/sower-jobs#29 in the environment's manifest. This job also has its own setup in cloud automation: uc-cdis/cloud-automation#1708 (#906) - 2 new fields in
gitops.json
:discoveryconfig.features.exportToWorkspace.[enableDownloadZip (bool) downloadZipButtonText (string)]
(#906)
- Update for PostgreSQL SSL support (#351)
- Add option for
isolation_level
READ_COMMITTED
. This addresses an error which occurs when there's competing transactions in thesheepdog
database, for example, when using multiple workers inuwsgi
which are trying to reach a shared resource insheepdog_db
. (#351) - The conflict occurs after submitting multiple requests to
sheepdog
using the gen3sdk for python (see https://github.com/uc-cdis/gen3sdk-python/blob/master/gen3/submission.py#L187) as part of populating the data dictionary with metadata. In this case, it looks like the race condition mentioned here https://github.com/uc-cdis/cloud-automation/blob/1f9cbb52cecf449a5ad76f984fe0dda449d924d3/kube/services/jobs/gdcdb-create-job.yaml#L48 is occurring. This is a workaround to useisolation_level
ofREAD_COMMITTED
withinsheepdog
. (#351)
- update to latest python base image (#356)
- Update
uwsgi.ini
andDockerfile
to supportpython
dependency installs (#351) - Update test suite (#351)
- Add debug log when unable to parse JSON input (#354)
- Handle TSV/CSV data for program/project creation (#354)
- Included
python
dependency updates inrequirements.txt
fordatamodelutils==1.0.0
,dictionaryutils==3.2.0
,gen3authz==0.4.0
,six>=1.15.0
,py2neo
, andgoogle-crc32==1.1.2
. (#351) -
Note that
dictionaryutils==3.4.0
is the currentsheepdog python
dependency but is not available for download from pypi (at least with awsl
/windows
release). The lower version3.2.0
is a workaround. (#351) - Included
python
dependency updates indev-requirements.txt
forcoverage==5.3.0
andcodacy-coverage==1.3.11
(#351)
- add filtering of list of jobs/outputs based on the user email (#23)
- Refactor and improve documentation: split the readme into several smaller
files and add
etlMapping.yaml
configuration details (#156) - Add example Azure DevOps pipeline and documentation (#151)
- This change will include parsing logic of uuid. (#33)
- Use env and config for populate script rather than passing parameters (#33)
- Add aggregate metadata adapters to talk to third party metadata services (#29)
- Better name for the defaulted index namespace (#33)
- Update documentation (#33)
- More verbose logging for populate (#31)
- Timeout retry via tenacity (#31)
- Remove aggregate MDS configs from source control (#30)
- A job has been created to make aggregate MDS population more robust uc-cdis/cloud-automation#1722 (#33)
- Must set the appropriate
AGG_MDS_NAMESPACE
value inmetadata.env
(#29) - Per #3332 a
metadata/aggregate_config.json
file should be created as part of the manifest config for a commons and the appropriate environment vars formetadata
should be defined in the manifest JSON (#30) - The new command to populate an aggregate MDS is
kubectl exec $(gen3 pod metadata) -- python /src/src/mds/populate.py --config /aggregate_config.json --hostname esproxy-service --port 9200
(#30)
- Bumps fastapi from 0.54.2 to 0.65.2. (#32)
- ES indices will not be namespaced per environment (#29)
- Scope ES indices to a namespace so they don't conflict with other environments (#29)
- This fixes an issue that prevents populating a AggMDS if one or more of the remote commons is down or unavailable. Instead of throwing an exception, the error is logged and the remote commons is skipped. (#31)
- Python package: tenacity (#31)
- gen3 job for syncing aggregate metadata to a metadata service (#1722)
- Added netnolimit for indexd (#1725)
- Added jenkins-ci cleanup to jenkins cronjob (#1712)
save-failed-pod-logs
now will save bothinitContainer
andcontainer
logs for specific pod (#1711)
- Increased header buffer size to prevent header too large error (#1724)
- Hatchery: mount
HOSTNAME
frommanifest-global
(#1717)