Hmmm,
Maybe it's time to resurrect the old crack rack. With Upgrades of course..... 😀
It's times like these I really appeciate the dnet clients, with the ability to install itself as a service, to buffer WUs, to tunnel firewalls, AND in this case to allow the client to fetch/flush WUs from a different directory (ie network drive) and the remote buffers option. 🙂
Although, now that I think about it,
Has anyone tried putting the clients on 1 network shared drive, mapping the network drive for each node, then running the SETI client from the node?
Let's see,
1. You would need 1 "server" node to run SETI, VNCVIEWER (or similar), a fast hard drive with more than enough disk space to hold all client's SETI folders.
2. Each node would have MB, CPU, RAM, FLOPPY, P/S, NIC
3. Node Operating System? Linux on floppy? Not knowing Linux, I'm assuming you can "map" a network drive similar to Windows OS.
4. Hub with enough ports for all nodes, server, and uplink to Internet connection or the rest of the network, cat-5 cables
5. One monitor, keyboard, mouse (or mouse on each, depending on if you can disable mouse in MB or OS) OR KVM switches if you don't want to be switching the cabling around, but after setup, it shouldn't be needed
6. For best uptime, a UPS(s) for the nodes
Then setup the "server" node, with SETIQ, VNCViewer, and the SETI directories. You could get each one started with a WU on the server, and have it configured the way you want.
SETUP a node SETI1, with VNCServer, NIC, hard coded IP address (or dchcp if network setup that way). Question is will Linux and VNCServer fit on 1 floppy? Could setup a RAM drive to run SETI from, since the hard drive access will be on the "server" hard drive and not the RAM drive. Execute SETI from that node's SERVER folder.
Each node would have to have it's own "startup floppy, with the IP address, the Network drive mapping set to it's own SETI folder on the server, then execute VNCServer, and SETI.
Once setup so the floppy will start the node automatically upon boot, shut down, unhook monitor, keyboard, etc., startup the node, check status on the "server" , then move on to the next one.
Obviously, quite a few details missing, but even if you had to use several floppies, once it was setup, restarts should be fairly quick once automated.
And, a lot of this is based on little or no knowledge of Linux, except for using Klinux, so I know at least some if not most of it is possible.
And I'm basically thinking "out loud" so to type. 😉😉
And, for the amount of time it would take me to figure out Linux and get it setup, it would probably be worth it to just grab a few cheapo hard drives.....😉😉😉