CPID different for Einstein project
Anonymous
1 Mar 2005 13:55:39 UTC
Topic 12785
(moderation:
For how long are you running the client that is attached to both (E@H and CPDN) now? Communication will take place via this one and may, in worst case, need two server contacts to both projects. So the time that it takes also depends on your settings how often to contact the server. Maybe you simply have to wait...
BM
CPID different for Einstein project
)
Zain is right, there are two entries in our database for your name. The one not related to your postings here is http://einstein.phys.uwm.edu//user/ . Both have different account keys, the one of the latter account beginning with a "d".
[Edit]: the email addresses are different. I'll drop a line to both.
BM
BM
Hm, unfortunatly there is no
)
Hm, unfortunatly there is no way to fiddle the other account out of the database without risking making it inconsistent. I'd suggest to at least change the name of the old account, but that's not really necessary.
Anyway I wonder how this could happen - I thought the names would be checked on craation?
BM
BM
> Someone can explain? I
)
> Someone can explain?
I haven't looked into that code yet, but there should be some flag on the client side that tells the server either "take this CPID for my account" or "I don't have a CPID yet - give me one". Apparently this flag somehow is broken on the Einstein-only machine, so every contact of each of the clients tells the server to take the CPID of the client. Maybe that's because the Einstein-only client was using the "wrong" account before and got a CPID for the wrong email address?
Try to stop the client on the Einstein-only PC, delete the line of the client_state.xml file (and maybe also from the sched_reply.xml, too), start the client again and see if the problem persists.
BM
BM
Just edited the posting - the
)
Just edited the posting - the brackets were lost...
BM
BM
This is a problem of BOINC,
)
This is a problem of BOINC, either the Core Client or the Client/Server communication. The Einstein@Home developers try to help the BOINC team if possible, but we curretly have our hands full with E@H specific stuff. I don't know enough of the BOINC code involved there myself, so if this will be fixed, then probably not by me. Also I think this would be hard to track down and doesn't affect many users anyway, so I doubt that has a high priority on the BOINC todo list.
BM
BM