Hide Forgot
Created attachment 1193457 [details] rhsm.log for 100 configure files Description of problem: We try to create 100 same configure files in /etc/virt-who.d for rhevm mode performance testing, that means virt-who will create 100 PIDs and fetch the hosts/guests info from RHEVM Server, some of the PIDs failed with "Unable to connect to RHEV-M server: 503 Server Error: Service Unavailable" Version-Release number of selected component (if applicable): virt-who-0.17-7.el7.noarch subscription-manager-1.17.10-1.el7.x86_64 How reproducible: always Steps to Reproduce: 1.create 100 configure files in /etc/virt-who.d for rhevm mode 2.run virt-who with oneshot option and check rhsm.log, many PIDs failed due to the following error: 2016-08-22 22:27:14,698 [virtwho.test-rhevm_1 ERROR] RhevM-1(8732):MainThread @virt.py:run:375 - Virt backend 'test-rhevm_1' fails with error: Unable to connect to RHEV-M server: 503 Server Error: Service Unavailable Actual results: some of the PIDs failed with "Unable to connect to RHEV-M server: 503 Server Error: Service Unavailable" Expected results: All the virt-who PIDs can fetch the host/guests info normally Additional info:
Given that virt-who connects to the same hypervisor 100 times in parallel, it's quite expected that one hypervisor won't be able to handle all the requests. Please retry with more than one hypervisor. As I see it now, this is not a bug.
one rhevm server can't handle so many virt-who PIDs at the same time, deploy 5 rhevm servers for testing, no error info found