I've started to update several hosts from 6.10.58 to 6.12.26 and a couple of them have exhibited oddball behavior.

Host ID "10986434" lost 17,827 credits the following report, and 14,428 the next day. It's pure cpu crunching... no GPU. It has a RAC of about 100.
http://boincstats.com/stats/boinc_host_graph.php?pr=bo&id=054e8fddeafd49b05cd2c717cec817ed

Thought it was a bug at first, but after reading the FAQ, it's probably a split CPID. I accidentally detached from Seti on this client and I think one more.

Is this behavior indicative of CPID splitting? If so, what's the timeframe for it to be updated back to correct info? I have Seti, Spinhenge, and Einstein running on about a dozen clients all with the same username, so the cross-client info is there to be salvaged from any of those other machines. I had intended to detach most clients from SETI and keep them running Spinhenge and Einstein because they are CPU crunchers only, and I wanted to get some GPUs set up for SETI.. Along the line of reprogramming this host and another (I think) I detached prior to crunching all the existing work units. I didn't realize it until it was done and tried to reattach, thereby causing the split CPID I assume.

Currently, this host is losing credit, and another host is reporting having no BAM data, despite it having some 500+ connections to BAM.

Reconnect with the hosts at all projects there were connected sometimes before and, if the host is show there as a new host, merge the computers by name. I had the same issue at some of my machines after upgrading from 6.12.26 to 6.12.33.

After the host has the new cpid at alls projects, detatch the machine from BAM, delete the old host in BAM Preferences and the attach the client newly.

This procedure works for me.

Regards Odi