Hi Rich, I had seen that bug but dismissed it as not likely relevant. It seems to be pretty specific to the import of a table. In our case, we had gotten past the table data each time we tried. We did find a way past it, but I do not at this point know which of the three things we did were the answer. Bah. 1) we set cluster_database to false 2) we increased parallel from 8 to 16 3) we excluded a view and a package that seemed like they might be at the root of the dependency chain - the other objects that we saw being compiled all reference these two objects. Some combination of those three got us past it. wish I knew which one, but no one felt we had the time to do one-at-a-time testing and figure out exactly which of these were our culprit. stephan _____________________________________________________________________________ Stephan Uzzell | MICROS Systems, Inc. Database Administrator - OPERA Global Technical Services 7031 Columbia Gateway Dr, Columbia, MD 21046 | 7 443.285.8000x2760 | 443.285.6505 -----Original Message----- From: oracle-l-bounce@xxxxxxxxxxxxx [mailto:oracle-l-bounce@xxxxxxxxxxxxx] On Behalf Of Rich Jesse Sent: Tuesday, 10 August, 2010 09:36 To: oracle-l@xxxxxxxxxxxxx Subject: Re: IMPDP blocking itself? Hey Stephan, > Looking at AWRs from any time during the time this has been stuck, the DP > workers all appear to be stuck on compiling various packages (I don't see > anything in v$sql or v$sqlarea with the sql_id's being reported in the AWRs > - but the module in AWR shows DP Worker) > > Any thoughts? > > We're on 10.2.0.4 on Win2k3 x64. Bug 7439689 - impdp worker process may spin (Doc ID 7439689.8) This bug looks to be specifically for Windows, although you'd have to trace it to know if it's looping in the same modules. GL! Rich -- //www.freelists.org/webpage/oracle-l