Status - 1. Continued work on HSZ failover design. A document (WITH fault containment section now ) is ready for review, though I've been considering that some of the HSZ support can readily be moved out of dkdriver code and into the server code, making the result less dependent on inter-layer cooperation. 2. Closed a few QARs. A mysterious data corruption problem in logged data was isolated and found to be happening AFTER DKdriver gets through with the data; it was reassigned to the error logger component. 3. Wrote slides for 3 DECUS talks; reviews will be scheduled for next week. 4. Working on some naming issues; there are several QARs on dual path naming problems which affect HSZ failover and also the planned extensions into failover to/from MSCP serving. I have now about exhausted the possibilities of finding any path to avoid devices appearing twice using existing code and flags...there are too many interdependencies so that no clean solution exists...and am dropping back to the idea of adding an extra flag to mark "extra" devices and suppress them from displays in most cases. 5. I continue to investigate the more global SCSI naming issues at a "background" rate, i.e., becoming more familiar with some of the boot path issues. Glenn Everhart