I am a newbie to Jabber. It seems a user can have numerous (possibly unlimited) resources and can send receive messages among its resources. If this is correct, for my very specialized client application that doesn’'t require authentication, it may be helpful for me to have a single user with a large number (thousands or more) of resources communicating among themselves.
Does it (wildfire) work the way I am describing?
Is the message passing performance among large number of users (with a single resource each) essentially same as a single user with large number of resources?
I wonder why foo@bar/a should send a message to foo@bar/b but this works with Wildfire. I didn’'t test it with 1000 resources but it should work fine (except that the cache is limited to 1000 identifiers, so you may want to patch the code or use only 998 different resources (foo and bar are the 2 to reach 1000).
the cache contains node, domain and resource identifiers (node@domain/resource) and it’‘s size is 1000. It’‘s used to verify the identifiers more quickly, so the “only” thing you’'ll see is a significant performance drop as the verification process (stringprep) takes long compared to the cache lookup.
I did recently create JM-855 to improve the cache as it has a little bug and to address the 1000 item issue.