Igor Mandrichenko

Nick, Have you verified that the rule gets updated ? Because if it does, I think that would contradict Stefan’s observations. Igor From: Nicholas Smith ***@***.***> Date: Friday, February 18,...

I have created a test case for this. See https://github.com/rucio/rucio/pull/5272

@mlassnig if I understand correctly, you are listing replicas in a dataset ? This is not the use case I am talking about. In my use case, the replicas are...

here is my code: ```python import sys, json from rucio.client.replicaclient import ReplicaClient files = json.load(open(sys.argv[1], "r")) print("loaded", len(files), "files") client = ReplicaClient() out = list(client.list_replicas([{"scope":f["namespace"], "name":f["name"], "did_type":"file"} for f in...

here is my modified code: ```python import sys, json, getopt, time from rucio.client.replicaclient import ReplicaClient opts, args = getopt.getopt(sys.argv[1:], "n:") opts = dict(opts) files = json.load(open(args[0], "r")) if "-n" in...

listing replicas by multiple file DIDs and by single dataset DID should make a difference because resolve_dids() has to resolve many file DIDs in the loop in first case while...

Looking at the code, I think it can be fixed in 2 different ways: 1. Let the "did_type" through to bypass the resolving DIDs loop 2. Resolve all DIDs in...

I will try to bypass the resolve_did step in my dev instance and see if that changes anything

I traced the time spent in various fragments of the server code participating in ``get_replicas()``. For 1000 files, in my smaller dev instance using a dedicated database with ~13K replicas,...

get_rse_name() should not be called 2000 times for a system with single RSE