Releases: ASFHyP3/hyp3
Releases · ASFHyP3/hyp3
HyP3 v10.16.0
HyP3 v10.15.0
Added
- Compute environments can now specify maximum vCPU, otherwise the default will be used.
Changed
- The
ItsLiveSpotIntelcompute environment now has a maximum vCPUs of 1000 to prevent ITS_LIVE production from overwhelming the STAC catalog.
Added
- Added an
INSAR_ISCE_BURST_EC2job type, that will use theDefaulton-demand compute environment. This is intended to be paired with theINSAR_ISCE_BURSTjob type defined in theINSAR_ISCE_BUSRT_SPOT.ymljob spec which uses theDefaultSpotcompute environment and allows end-users to submit either spot or on-demand jobs. - Added the
INSAR_ISCE_BURST_EC2job type to the Cargill custom HyP3 deployment.
Fixed
- Reverted the
r6id[n]upgrade tor8idinstances for spot compute environments due to lack of spot capacity. This includes theDefaultandDefaultSpotenvironments forhyp3-cargilland theItsLiveIntelandItsLiveSpotIntelcompute environments.
HyP3 v10.14.5
Changed
- Reduced default/expanded vCPUs in DAAC deployments to 600/1200 to account for pricing changes introduced in v10.14.2 and v10.14.3.
HyP3 v10.14.4
Changed
- Upgraded custom deployments from the
r6id/r6idnandm6id/m6idninstance families to ther8idandm8idfamily. - The AMI IDs specified in
compute_environments.ymlare now resolved by CloudFormation instead of hard-coded. - Simplified the ITS_LIVE custom compute environments.
HyP3 v10.14.3
Changed
- Upgraded DAAC and Plus deployments from the
r6id/r6idninstance families to ther8idfamily.
HyP3 v10.14.2
Changed
- Updated all job specs in all deployments to use On Demand instances except for:
AUTORIFTjobs in hyp3-its-live and hyp3-its-live-testINSAR_ISCE_BURSTjobs in hyp3-cargill
Removed
- Removed support for
OPERA_RTC_S1jobs:OPERA_RTC_S1job type from edc-uat and edc-sandbox environmentsOperaRtcS1EndDatestack parametercheck_opera_rtc_s1_boundsandcheck_opera_rtc_s1_datevalidators and associated tests
HyP3 v10.14.1
Changed
- PISM job specs no longer need to provide an RGI geopackage.
HyP3 v10.14.0
Added
- Job specs may now specify any LinuxParameters using the
linux_parameterskey on a per-step basis. - Added three new PISM job specs to prepare and run ensemble simulations of any RGI v7.0 glacier complex to the PISM-Cloud deployments.
- Added a new
VOLCSARVATORY_MINTPYjob spec which pulls multiburst products and generates time series files using MintPy. - Added support for NISAR RSLC and GSLC scenes to the
ITS_LIVE_AUTORIFTandARIA_AUTORIFTjob specs.
Changed
- Implemented dynamic memory allocation for multi-frame
AUTORIFTjobs inset_batch_overrides.py. Sentinel-2 jobs with more than 4 frames now use 16 GB of memory instead of 8 GB. Single-frame Sentinel-2/Landsat and Sentinel-1 memory allocation is unchanged, while NISAR memory allocation is 64 GB. - Increased the timeout from 3 hours to 6 hours on the hyp3-autorift step in the
ITS_LIVE_AUTORIFTandARIA_AUTORIFTjob specs to account for NISAR runtimes. - Updated
OPERA_DIST_S1job spec to allow for entrypoint change (now need to specifyrunat beginning of command). - Added
OPERA_DIST_S1_CONFIRMATIONfor confirming products OPERA DIST-S1 Products. - The
SRG_GSLCandSRG_TIME_SERIESnow allow processing Sentinel-1 C and D scenes.
Fixed
- Resolves DIST_S1_CONFIRMATION issue related to length of inputs; see #3041
HyP3 v10.13.1
Fixed
- Upgraded to cryptography v46.0.5 from v46.0.3. Resolves CVE-2026-27006.
HyP3 v10.13.0
Added
- Added optional
chip_sizeandsearch_rangeparameters to theARIA_AUTORIFT.ymljob specification to enable user-definedchip-sizeandsearch-range. - Updated
AUTORIFT.yml,ARIA_AUTORIFT.yml, andITS_LIVE_AUTORIFT.ymlvalidation schema to support processing of Sentinel-1D and Sentinel-2C/D granules. - Added a
model_context_lengthparameter to theOPERA_DIST_S1job specification. - Added a new custom hyp3-slimsar-test deployment.
- Added a new
SLIMSAR_TDBPjob_spec for slimsar time-domain backprojection processing. - Added a new
SlimSARcompute environment for slimsar processing with correct EC2 instances. - Added a new
ITS_LIVE_CROP_BULKjob spec which re-crops existing ITS_LIVE products which are specified in a parquet file to ensure they are chunk-aligned and have a time dimension, and then it generates STAC JSON and other metadata files. - Added a new
ITS_LIVE_META_BULKjob spec which generates STAC JSON and other metadata files for existing ITS_LIVE products which are specified in a parquet file. - Added the
ITS_LIVE_CROP_BULKandITS_LIVE_META_BULKjob spec to the ITS_LIVE deployments. - Added the
stac_items_endpointandstac_exists_okayjob parameters to theITS_LIVE_AUTORIFTandITS_LIVE_CROPjob specs to allow directly publishing STAC items to the ITS_LIVE STAC catalog.
Changed
- Increased the maximum
stride_for_norm_param_estimationto 32 from 16 for theOPERA_DIST_S1job specification to handle models with a larger input size. - Changed readme to reflect current state of AWS, motivation for deploying a hyp3 stack, and clarify naming and formatting.
- The
ITS_LIVE_METAjob spec has been renamedITS_LIVE_CROPas it re-crops an existing ITS_LIVE product to ensure it is chunk-aligned and has a time dimension, and then it generates STAC JSON and other metadata files. - The hyp3-ci stack permission for JPL deployments were expanded to support deploying HyP3-based monitoring stacks:
- Listing CloudFormation stacks is now allowed.
- CloudFormation permissions were expanded to any region from just us-west-2.
- ECR actions are now allowed.
Removed
- Removed the
publish_stac_prefixjob parameter from theITS_LIVE_AUTORIFTandITS_LIVE_CROP(previouslyITS_LIVE_META) job specs as it's no longer used by the ITS_LIVE plugins.