>> Maybe there are some hints we could use.. > > Useful, but not revolutionary.
Never claimed it was > But yes, a VM based store has it's benefits, but for that to work for us > a different design is needed more along the lines of Varnish, making > heavy use of worker threads for mostly every operation, not just one per > core. Actually the thing I found the most interesting is that it suggests to use page-aware object placements so that big structures traversal is easier on the VM. Could it be useful to adopt that for some of our low-level indexes? We do have a few hashes and trees laying around which maybe could benefit from this; and adopting an alternate algorithm for those may not have a big impact code-wise.. -- /kinkie