Xen Io isolation method and CPU scheduling script

IO:
1. Virtual Disk I/O management mechanisms in xenhttp: // lists.xensource.com/archives/html/xen-devel/2007-07/msg00863.html
2. Token-based-QoS-resource-limiting-for-vbd-I-ohttp: // token
3. DM-iobandhttp: // sourceforge.net/#/trac/ioband/wiki/dm-ioband/man/examples
4. ionicefor blkpid in $ (pgrep blkback); do # 2nd lowest (6) PRIO in best-effort class (2) ionice-C 2-N 6-p $ blkpiddone



CPU:
#! /Bin/sh
# Cpucap # wartungsfenster.de # reschedule all running xen Domus to a medium-low CPU allocation # scheduler notes: # If you wanna change this to support QoS classes (I. e. if the customer can # chose a certain performance level, then look at the sedf and credit2 # schedulers, or just play around with the weight. # I recommend using the weight for now. # also I think a 2-core cap + very heavy weight fo R dom0 shoshould be looked # at when you see any Io scalability issues. # hyperthreading notes: # We might see hyperthreads so we'll assume 2 cores have a # performance of 130 => and 1 core has 65! # This means we'll have to tweak xen vcpu allocation to ensure noone ever # runs on a HT only? # Potential solution for figuring out # The best solution is using a extended CPU mask (URL ?) In the CPUs = line # Of the xen config. that, coupled with picking specific core types # (URL: http://dag.wieers.com/blog/is-hyper-threading-enabled-on-a-linux-system), wocould work, but is extremely complex. # I will focus on giving the users multiples of 1 core + 1ht and stick by the #130 sum then... get_vms () {XM list | egrep-V 'name | Domain-0' | awk '{print $1 "" $2 "" $4}'} set_perf () {XM sched-credit-D domain-ID-W Weight = C cap} # xen4.1/sched-credit2 testers needed: # The below command is disabled for safety reasons, can crash dom0 # http://lists.xensource.com/archives/html/xen-devel/2011-03/msg01762.html # I bet theres still no regression test for that. # Our first try, later we shoshould give dom0 256 * cores * total_domus # Or just pin it, if we got lotsa fast cores default_weight = 256dom0_cores = $ (grep-C ^ proc/cpuinfo) dom0_weight = $ ($ default_weight * $ dom0_cores )) # XM sched-credit2-D 0 weight $ dom0_weight # We shoshould employ capping for dom0 too, feel free to move to the main code. # XM sched-credit2-D 0-C 400 echo "$ (get_vms)" | while read name domid vcpus; do XM sched-credit-d $ domid-C $ ($ vcpus * 65) done