-
Story
-
Resolution: Done
-
Major
-
DO280 - OCP 3.4 1 20170330
-
None
-
en-US (English)
URL:
Reporter RHNID:
Section: -
Language: en-US (English)
Workaround:
Description: At random times during the course, even when performing labs that have nothing to do with it, the Nexus running on workstation.lab.example.com consumes an important amount of compute resources, thus blocking any other action from the user on the machine.
This is the VM from which students are intended to work. Therefore some labs can be tedious because of the system becoming unresponsive.
As discussed on the GLS Contractor Mailing List (Message-ID : CAOPSvMQeXOhTeJFMu-kfyOrxS4WNdqeQvApQ_pjgiYfKXBnnsg@mail.gmail.com), this could be managed by either :
- Using a Linux Control Group to limit Nexus' compute resource consumption
- Try to adjust Nexus' configuration and JVM tunables to have it behave more efficiently and take down any unneeded functionality
- Upgrade the memory and CPU characteristics of workstation
- Isolate Nexus in a separate VM for a finer control over its limits and the ability to reset any one of workstation and this new VM without perturbing the other
Or a combination of all four.