the only reasons I can think about why not to use 64bit Kernel support are:
- You need a machine that really takes advantage of the technology. With or above 4GB RAM you should switch to 64Bit.
- 64Bit Kernel support is rather new and there are still some unknown issues to be dscovered and fixed in the Linux Kernel and tools
- At the ASG there are very few (if any) daemons that already take advantage of 64Bit. The main advantage is increased allocateable memory (increased max connections etc.). It's only the Kernel that makes the difference at the ASG as of now. You can also easily switch kernels by (un)/installing the corresponding kernel RPMs.
No words about performance from me ;-). Not that I wouldn' t tell you, just a lack of the necessary figures.
1.) Obviously some benchmarks could be very helpful to the decision making process. The justification of a move to 64 bit almost requires it in a business environment. Particularly in light of any risk of unknown issues or problems with 64 bit kernels in an ASG. Let's call it necessary additional motivation.
2.) What kind of timeline would you think one could expect with regard to the implementation of 64 bit daemons for most, if not all, of the various processes?
3.) And regardless of the timeline, what are your thoughts about the performance improvements to be had via 64 bit daemon conversions.
1.) Obviously some benchmarks could be very helpful to the decision making process. The justification of a move to 64 bit almost requires it in a business environment. Particularly in light of any risk of unknown issues or problems with 64 bit kernels in an ASG. Let's call it necessary additional motivation.
2.) What kind of timeline would you think one could expect with regard to the implementation of 64 bit daemons for most, if not all, of the various processes?
3.) And regardless of the timeline, what are your thoughts about the performance improvements to be had via 64 bit daemon conversions.
