|
HP OpenVMS Systemsask the wizard |
|
The Question is: We are having a strange problem. We have a Vaxcluster consisting of a pair of VAX 6000/440's and a single Vaxstation 3100. Our disk drives that are available to the Cluster consist of about 19 two megabyte MTI Stingray 220's that are served by four disk servers. We've had them for 8 or 9 years and experienced no major problems with operations or replacement issues. We've usually popped the old drive out and replaced it with a new. One of the drives failed horribly last week, and we called our regular MTI technician in to replace it. This time however we've not been able to initialize the new drive, because the drive always seems to be allocated by the other Vax when it boots. We've commented it out of our startup procedure, but the booting system always seems to grab it, or the non booting system fails to see it as released if the other system is not present. We also tried booting the alternate system with a minimal conversational boot so it doesn't hold it, despite this, the other system still saw it as allocated. When we tried to boot this system without the other members of the cluster, it then hung on startup and the console froze as the original system was doing. This is an unpleasant thing, for some reason whichever console is booting tries to access it, and then hangs. When we try kill the opcom terminal process or stop opcom itself, nothing happens. When I try to sho proc/cont/id= on the opcom terminal that has the disk drive allocated comes up as a suspended process, even though a sho sys returns it as LEF. Either way it sure is hung, and worse does so is the process after it allocates the drive. We can't figure out why it's trying to allocate (successfully) the drive since we commented it out in the startup file. Powering the drive up or down doesn't seem to do anything useful. The vaxstation is usually out of the picture. Any ideas, I just took over this job, and this is killing me. The MTI technician and my predecessor have never seen this before and are stumped. We want to avoid losing a drive if possible as we're pretty maxxed out her. Thanks, Don Perreault Northeastern University The Answer is : Assuming that this is OpenVMS involved and not something specific to how the MTI drives connect to OpenVMS or in how they present a host interface to OpenVMS, shut down all cluster member nodes and reset or otherwise INIT the processors, boot exactly one of the cluster nodes (one with direct access to the MTI drive) minimally, and then see if you can access the drive. Please contact MTI for assistance with any considerations specific to the MTI storage devices. If the MTI service organization is unable to resolve this, please contact the Compaq Customer Support Center for further assistance in troubleshooting this.
|