That's always in demand, and probably more after the apocalypse. Whiskey and ammunition...
Slashdot videos: Now with more Slashdot!
HorizSync 28.0 - 33.0
VertRefresh 43.0 - 72.0
VendorName "NVIDIA Corporation"
Option "SLI" "1"
Option "Coolbits" "4"
AH.... thanks for asking. I forgot teh "SLI" part : )
The NVS450 has 4 outputs; I'm using three of them. And they're not in the right order on the desk (according to the card)
I should mention I'm using relatively new drivers from NVidia's site, not the apt-able ones. 313.09, specifically, though after the 310.x series BaseMosaic hasn't broken any more.
I can confirm that BaseMosaic on an NVS450 works under LMDE (Debian Testing) using:
Option "BaseMosaic" "True"
Option "MetaModes" "GPU-1.DFP-0: 1680x1050+0+0, GPU-0.DFP-1: 1680x1050+3360+0, GPU-0.DFP-0: 1680x1050+1680+0; GPU-1.DFP-0: NULL, GPU-0.DFP-1: NULL, GPU-0.DFP-0: 1680x1050"
This is such a ridiculous comment I had to actually reply, and that doesn't happen often. I would have dismissed it as a troll, but I think you're serious. I'm a DBA (mostly MySQL + random stuff like DB2, Mongo, etc) and we're heavily virtualized on real workloads, real 24x7, on a product you've definitely heard of. And we're not incompetent. Doing real virtualization (we use VMWare with VSphere) is fantastic because:
1) Moving VMs between hosts with no downtime.
2) Hardware abstraction layer
2.a) Hardware upgrades with no downtime to any service
2.b) VM failover on the fly
2.c) Move VMs between datacenters
3) Cloned spinup
4) Snapshot backups (with OS integration)
5) On-the-fly storage expansion
6) multi-SAN connectivity
7) Resource pooling
8) Cost effectiveness
9) Resource oversubscribe (production, but typically unimportant machines get things like the memory balloon driver)
10) Rebalance of resources as workloads change.
Where virtualization really sucks (at least on VMWare)
1) SMP/multi-core VMs
2) Purple Screen of Death
3) 2TB limit on LUN size on ESX 4.x