Age | Commit message (Collapse) | Author | Files | Lines |
|
for collectd KPIs we need may need to open 10,000+ files
one file for each core * number of CPU counters we read.
Use the ansible pam_limits moules to increase the nofile
setting. This should allow the VNF to boot with increased
number of files open
Change-Id: I46b75d5ca59e8e359c985a1e37daaee89897884c
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>
|
|
Added review dependency, removed collectd commit-id
replicating bash script, added pmu roles to playbook
Create local mirror for event list generation, copy script
to image, and run event list generation at runtime if
intel_pmu is enabled.
Change-Id: Ie46a2b197f4d2037cf3eed194764ce9eb5670415
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>
|
|
libyajl needed for ovs_events and ovs_stats, removed installing
from src and rather installing as deb pkg
Change-Id: Ic1ebf88304bbc7f4fa7b9b51c95f1e8e729e71e3
Signed-off-by: Dhaval Patel <dhaval.r.patel@intel.com>
|
|
required DPDK fixes to enable building collectd with dpdkstats
need /usr/lib/dpdk-pmd symlinks
Compiling all this code requies more space than the default ubuntu
cloud-image, so we need to resize the cloud image somehow
Change-Id: Ib39fc94ba8101ca1aebc7d2a710be88b0338f7c3
Signed-off-by: Ross Brattain <ross.b.brattain@intel.com>
Signed-off-by: Maciej Skrocki <maciej.skrocki@intel.com>
|