Open eduard93 opened 4 years ago
I would like to clarify that standalone IRIS images start successfully.
Also WSL does not impose CPU/Memory limits the way old HV backend did.
Hi! I just upgraded my windows box to use WSL 2. I am using the same version of docker as you. The only difference is that I picked Ubuntu 18.04 instead of 20.
Mine is working fine.
Maybe you are having this problem because you are using a version of Ubuntu that is not supported by InterSystems IRIS 2019.3 (the version of IRIS used by this demo)?
We are going to upgrade to IRIS 2020. We are just waiting for a more stable build.
Kind regards, AS
How's your RAM consumption?
Moved to U18, things got better (no crashes) but now vmmem consumes all free memory ~18Gb and hangs the host.
Testing this workaround.
I will take a look. Mine was running ok. IRIS is limited to take just a portion of memory for global buffers. It is not automatic memory allocation. The Spark cluster was also limited, I guess to 1G per worker plus 1Gb for the master. The UI + the two IRIS boxes and the Zeppelin node should be taking just a little more than 1Gb of RAM. Do you know who is taking. The demo should not take more than 4Gb of RAM total.
Can you look what process is consuming most of the memory?
I will try again on my PC later today...
Workaround works. I'm running only this demo with RAM limit at 8 Gb and RAM consumption grows to ~8Gb and stops after that.
Without 8 Gb limit the consumption grows further.
I'm unable to start the demo with DD on WSL2 backend. Here's the error info. How can I fix that?
Versions:
I'm integrating WSL with Ubuntu-20.04 image
Ubuntu has free space (other images do start):
Got these errors on IRIS containers: