Bug: list-datasets-rse depends on COLLECTION_REPLICAS which we stopped populating properly

[haozturk]

Bug Description

list-datasets-rse depends on COLLECTION_REPLICAS [1] which we stopped populating properly with #864 Site admins use this command often. Max made me realize it at [2]

[1] https://github.com/rucio/rucio/blob/d9a818c1c07c1aa143a3880e6350fbece1640ecf/lib/rucio/core/replica.py#L3526
[2] https://its.cern.ch/jira/browse/CMSDM-234?focusedId=6642819&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-6642819

Reproduction Steps

$ rucio list-datasets-rse T2_US_MIT_Tape | grep cms:/ADDGravToGG_NegInt-0_LambdaT-4000_M-2000To3000_TuneCUEP8M1_13TeV-pythia8/RunIISummer16MiniAODv3-PUMoriond17_94X_mcRun2_asymptotic_v3-v1/MINIAODSIM
cms:/ADDGravToGG_NegInt-0_LambdaT-4000_M-2000To3000_TuneCUEP8M1_13TeV-pythia8/RunIISummer16MiniAODv3-PUMoriond17_94X_mcRun2_asymptotic_v3-v1/MINIAODSIM#93687308-a1e4-4b40-8d4b-a1ab09c1bf0c
cms:/ADDGravToGG_NegInt-0_LambdaT-4000_M-2000To3000_TuneCUEP8M1_13TeV-pythia8/RunIISummer16MiniAODv3-PUMoriond17_94X_mcRun2_asymptotic_v3-v1/MINIAODSIM#99f1360e-de4a-46b7-a743-5e5fc89f29e9
cms:/ADDGravToGG_NegInt-0_LambdaT-4000_M-2000To3000_TuneCUEP8M1_13TeV-pythia8/RunIISummer16MiniAODv3-PUMoriond17_94X_mcRun2_asymptotic_v3-v1/MINIAODSIM#cf8a0a24-2620-45db-a453-6afe34d82915
cms:/ADDGravToGG_NegInt-0_LambdaT-4000_M-2000To3000_TuneCUEP8M1_13TeV-pythia8/RunIISummer16MiniAODv3-PUMoriond17_94X_mcRun2_asymptotic_v3-v1/MINIAODSIM#e30107fd-dd7e-4b66-9b60-b3df3bb34435

$ rucio list-dataset-replicas cms:/ADDGravToGG_NegInt-0_LambdaT-4000_M-2000To3000_TuneCUEP8M1_13TeV-pythia8/RunIISummer16MiniAODv3-PUMoriond17_94X_mcRun2_asymptotic_v3-v1/MINIAODSIM#e30107fd-dd7e-4b66-9b60-b3df3bb34435

DATASET: cms:/ADDGravToGG_NegInt-0_LambdaT-4000_M-2000To3000_TuneCUEP8M1_13TeV-pythia8/RunIISummer16MiniAODv3-PUMoriond17_94X_mcRun2_asymptotic_v3-v1/MINIAODSIM#e30107fd-dd7e-4b66-9b60-b3df3bb34435
+-----------------+---------+---------+
| RSE             |   FOUND |   TOTAL |
|-----------------+---------+---------|
| T2_US_Nebraska  |       1 |       1 |
| T2_IT_Legnaro   |       1 |       1 |
| T0_CH_CERN_Tape |       1 |       1 |
+-----------------+---------+---------+

list-dataset-replicas shows the correct result

Expected Behavior

Well, this command should reflect the correct result

Possible Solution

We can do a patch similar to what we did to list-dataset-replicas [1] which will make the server calculate the result on demand. But this would be a very expensive calculation. I don't think it'd work.

[1] https://github.com/dmwm/CMSRucio/blob/master/docker/rucio-server/Dockerfile#L36C5-L36C76

Related Issues

@ericvaandering @Panos512 @eachristgr FYI