As you might have noticed CFD Online has been off-line a few times over the last two weeks. We have suffered several hard-disk failures in our redundant RAID5 disk array. We replaced the broken disks two times, but when the problems occured a third time on the new disks we gave up and instead ordered a brand new server.

CFD Online is now running on this new server. Hopefully the problems will now be history. The disk crashes have caused some data-loss. We apologize about this and you are very welcome to resubmit data that is missing.

The new server is a bit faster than our old server. Hopefully this will improve response times for those of you who have very fast network connections. If you are interested in what server we have now, CFD Online is now hosted on a dedicated:

HP ProLiant DL380G4
2 x Intel Xeon 3.6 GHz CPUs with 2MB cache and 64-bit capabilities
4 GB (4 x 1024 MB) of 2-way PC2-3200R 400MHz DDR2 RAM memory
Ultra320 Smart Array 6i RAID Controller
Hot Plug Fully Redundant Fans
Hot Plug Redundant Power Supplies
Four 73GB 15krpm U320 SCSI disks in a RAID5 array with an online spare disk
Two 300GB 10krpm U320 SCSI disks mirrored for backups



The server is located in a datacenter in Salt Lake City, Utah, USA.

Our webhost, who has done a tremendous job to help us over the last weeks, is www.servmatix.com

The question is that checking cfd-online has become a daily habit just like your morning coffee Wish you all the best

Would be nice if the hosting service offered the ability to reroute (if the server went down) all user page requests (including previously bookmarked pages deep in the cfd-online website hierarchy) to an "emergency webpage" stored on a different server (perhaps one server the hosting service sets aside for emergency pages of all their customers). The emergency webpage, which ideally should be editable by the particular customer or the hosting service to enter ETA for restored functionality, would carry a brief message about technical difficulties and when they might be overcome. Would save site users from having to repeatedly kill their hung browser process.