I had the same problem with the kubernetes lemmy.ml community yesterday night, but this morning I was able to find it through the search and subscribe to it (although for the moment it seems quite calm to say the least :)). I'm probably wrong here, but the lemmy.ml staff posted an announcement stating it was quite overloaded with the influx of new users - maybe to keep things afloat they are rate-limiting some resource-hungry api calls?

edit: also, as @anthonyg just mentioned, it might also be a propagation time between instances - if I understand correctly the sdf lemmy instance is quite recent.