You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
nadove-ucsc opened this issue
Sep 28, 2024
· 0 comments
Assignees
Labels
-[priority] Mediumbug[type] A defect preventing use of the system as specifiedindexer[subject] The indexer part of Azulorange[process] Done by the Azul team
The linked issue adds replicas that were previously missing for many HCA entities, such as donors and some protocols. However, there are still HCA entities that are not being replicated. There are two distinct cases:
Entities that are linked to a file, but are not replicated because they are not tracked while traversing the links. An example is the dissociation_protocol in canned bundle aaa96233-bf27-44c7-82df-b4dc15ad4d9d.
Entities that are not linked to any file in their bundle.
The solution for case 1 is to modify the TransformerVisitor class to track all linked entities it encounters, potentially consolidating all currently untracked entities in a single data structure. These entities will then be emitted as replicas by the FileTransformer.
The solution for case 2 is to modify the ProjectTransformer to emit a replica for every entity in its bundle. The hub IDs for these replicas will not include any file IDs. Duplicate replicas will be merged by the index service before any replicas are written to ElasticSearch.
This design depends on the current implementation of the linked ticket, as in #6584
The text was updated successfully, but these errors were encountered:
nadove-ucsc
added
orange
[process] Done by the Azul team
bug
[type] A defect preventing use of the system as specified
indexer
[subject] The indexer part of Azul
-
[priority] Medium
labels
Sep 28, 2024
-[priority] Mediumbug[type] A defect preventing use of the system as specifiedindexer[subject] The indexer part of Azulorange[process] Done by the Azul team
Follow-up from #6582
The linked issue adds replicas that were previously missing for many HCA entities, such as donors and some protocols. However, there are still HCA entities that are not being replicated. There are two distinct cases:
dissociation_protocol
in canned bundleaaa96233-bf27-44c7-82df-b4dc15ad4d9d
.The solution for case 1 is to modify the
TransformerVisitor
class to track all linked entities it encounters, potentially consolidating all currently untracked entities in a single data structure. These entities will then be emitted as replicas by theFileTransformer
.The solution for case 2 is to modify the
ProjectTransformer
to emit a replica for every entity in its bundle. The hub IDs for these replicas will not include any file IDs. Duplicate replicas will be merged by the index service before any replicas are written to ElasticSearch.This design depends on the current implementation of the linked ticket, as in #6584
The text was updated successfully, but these errors were encountered: