follow on to #118, registry-manager still unable to change archive status on bundle contents #136
Comments
|
@jimmie @tloubrieu-jpl @alexdunnjpl can we talk about this at the breakout today? do we need to update the SBN registry to make this work? |
jordanpadams
changed the title
|
@jordanpadams not sure I understand - this is a duplicate of #118 (and registry-common #29 and registry-mgr #57), resolved in registry-common #30 |
|
Regarding the existing data in the SBN registry, it will be necessary to re-harvest affected bundles with the |
|
@alexdunnjpl sorry for confusion. the re-ingesting part was what I didn't get from the other tickets. @plawton-umd can you try re-ingesting the bundle with the |
|
@plawton-umd in case it isn't clear, the re-ingestion will only work when using an updated version of the software. |
|
@jordanpadams @alexdunnjpl If I understand correctly, the SBN registry needs to be updated to the new version of the registry software (in addition to the changes needed to deal with the shards issue). Once that registry software update is done, then I should re-ingest the data and attempt to update the archive status. When is the EN controlled SBN UMD's registry's software being updated? (Note: My availability for the next 3 days is limited, so next week can work - I am not attending AGU.) |
|
@plawton-umd sorry for the confusion. we do not believe anything needs to be updated on our end. since you have the updated registry-mgr, you should just need to install the latest snapshot of harvest, and re-harvest / re-ingest the bundle that you are having difficulty with. |
|
apologies for the inconvenience. |
|
@jordanpadams Installed harvest-3.8.0-SNAPSHOT. Tried -O and a different time tried --overwrite. Same results - summary below. Via open search the products are still from the Oct 2022 load. ./harvest-3.8.0-SNAPSHOT/bin/harvest -O -o epoxi_mri/out -l epoxi_mri/harvest.log -c epoxi_mri/epoxi_mri.cfg Tried registry-manager anyway .... same error ./registry-manager-4.6.0-SNAPSHOT/bin/registry-manager set-archive-status -status archived -lidvid "urn:nasa:pds:epoxi_mri::1.0" -es $OPENSEARCH_URL -auth $AUTH_FILE -l mri_20221206c_status.log Not sure what I am doing wrong. |
|
@plawton-umd it's unclear to me why harvest is ignoring Could I trouble you to drop a copy of your @jordanpadams @tloubrieu-jpl are there any differences that jump out at you with me running the main class directly from my IDE (see first line) vs the as-installed CLI command? I may have to do a proper installation from the snapshot to verify 1:1. |
|
@plawton-umd @alexdunnjpl no idea. i would recommend installing and running from CLI and see what happens sorry again @plawton-umd ! we will get to the bottom of this. thanks for being patient. @tloubrieu-jpl @jimmie any chance you have some bandwidth to help test this bug and figure out what is happening here? |
|
@jordanpadams @plawton-umd I will look at that tomorrow. |
|
@plawton-umd I tried the --overwrite option on one f my harvest configuration file and that worked (it did not skip any file). But I guess my configuration file is different from your and there might be a case that we overlooked while testing the --overwrite option. Were you able to send your harvest configuration to @alexdunnjpl , if not can you send it to me ? or just attach it to this ticket (drag and drop works). Note that until you are not able to reload the file in the registry (with harvest) you don't need to test the registry-mgr again since the error came originally from the way harvest loads the data in the registry. |
|
@plawton-umd , @alexdunnjpl shared your configuration file with me. The harvest/overwrite issue happens when you use the tag So a work-around for this ticket would be for you to use We will work on the bundles/overwrite issue in a different ticket on harvest. See NASA-PDS/harvest#111 Sorry that you have to experience all these bugs. |
|
Fixed in NASA-PDS/harvest #113 |
🐛 Describe the bug
When changing a bundle's archive status to archives, the following error was displayed
[ERROR] Expected BEGIN_ARRAY but was STRING at line 1 column 166 path $._source.ref_lid_collection
This is a reopen of #118
📜 To Reproduce
See #118
Set OPENSEARCH_URL to
https://search-sbnumd-prod-o5i2rnn265gnwmv2quk4n7uram.us-west-2.es.amazonaws.com:443
or equivalent
Set AUTH_FILE to your authorization credentials file
Run a command similar to
./registry-manager-4.6.0-SNAPSHOT/bin/registry-manager set-archive-status -status archived -lidvid "urn:nasa:pds:epoxi_mri::1.0" -es $OPENSEARCH_URL -auth $AUTH_FILE -l change_status.log
[ERROR] Expected BEGIN_ARRAY but was STRING at line 1 column 166 path $._source.ref_lid_collection
🕵️ Expected behavior
I expect the ops:Tracking_Meta/ops:archive_status to be changed from "staged" to "archived"
for all collections and products in the bundle.
📚 Version of Software Used
./registry-manager-4.6.0-SNAPSHOT/bin/registry-manager -V
Registry Manager version: 4.6.0-SNAPSHOT
Build time: 2022-12-02T17:06:38Z
🩺 Test Data / Additional context
The log file contents were
2022-12-06 11:54:11,404 [INFO] Setting product status. LIDVID = urn:nasa:pds:epo
xi_mri::1.0, status = archived
The screen messages were
[INFO] Setting product status. LIDVID = urn:nasa:pds:epoxi_mri::1.0, status = archived
[ERROR] Expected BEGIN_ARRAY but was STRING at line 1 column 166 path $._source.ref_lid_collection
🏞Screenshots
🖥 System Info
OS: [e.g. iOS]
Linux
Browser [e.g. chrome, safari]
n/a
Version [e.g. 22]
🦄 Related requirements
⚙️ Engineering Details
The text was updated successfully, but these errors were encountered: