Issue
I deployed openstack packstack (all-in-one) in centos7 server with 8vcpu and 16gb memory. Deployment of packstack openstack successful completed. I can spin instances with flavor with 2gb or less but when use 4g or 6g ith will return fail instantiate error.
Exceeded maximum number of retries. Exhausted all hosts available for retrying build failures for instance
This is clean openstack deployment on clean centos7 server.
when I run free -m it show below
total used free shared buff/cache available
Mem: 15845 14816 259 532 769 197
Swap: 8063 5049 3014
This is the 1st VM to instantiate..I puzzle and don't know where all the memory went.
From the hypervisor view
Type : QEMU
VCPUs (used): 0
VCPUs (total): 8
RAM (used): 0
RAM (total): 15.5G
Local Storage (used): 0
Local Storage (total): 49GB
Instances: 0
I have seen people with similar problem without any solution...I really need help from all you guyz...Hope someone could lead me the way.. Please do help ..Thank you
Solution
In my experience, recent versions of Packstack are rather memory-hungry. You can use top and sort by VIRT and/or RES to get an idea which components use memory most. Then try switching off parts that you don't need.
On your running server, you could try stopping the corresponding systemd services if you don't need them. You could also look for parameters in /etc/<service>/<service>.conf that configure number of workers, reduce them and restart the service. Example: osapi_compute_workers in one of the /etc/nova config files.
Alternatively, generate a default answer file and set appropriate parameters like CONFIG_something_INSTALL to "n". I don't know what Packstack installs by default these days; CONFIG_HEAT_INSTALL or CONFIG_CEILOMETER_INSTALL are hot candidates (of course, only if you don't need them). Then reinstall from scratch.
Answered By - berndbausch Answer Checked By - Cary Denson (WPSolving Admin)