question MAP/REDUCE stuck at 0% in Archives of Support Questions (Read Only)

MAP/REDUCE stuck at 0%

oninom — Fri, 16 Sep 2022 10:34:37 GMT

I have problem with fresh installation on pseudodistributed Cloudera manager. I have one VM with 12GB ram and 4 cores. All examples are stuck at 0% and are not going. In http://node1.bigdata:8088/cluster/scheduler# Memory used is always 0 same is for cores. Any idea what I can do to fix this.

Edit: I checked Cloudera QuickStart and in :8088 I have alocated 8GB of memory while in my fresh installation only 1GB how I can change this. I know that most issues with this are about memory but I'm unable to find proper solution and guide for file/settings location. Hadoop installed with CM store setting in different location that installed normally.

Re: MAP/REDUCE stuck at 0%

oninom — Fri, 12 Aug 2016 18:45:06 GMT

Ok, I managed to fix this. I used https://www.cloudera.com/documentation/enterprise/5-7-x/topics/cdh_ig_yarn_tuning.html and downloaded spreadsheet file. I used values from this spread sheet and put it in my config. Becouse I used CM for install to config this its necesary to go to http://ip:7180/cmf/home click on Yarn(M2 included). Then click on configuration. In search bar use yarn.nodemanager.resource.memory-mb and change it for around 4-8 times more than yarn.scheduler.minimum-allocation-mb. I put in yarn.scheduler.minimum-allocation-mb 1024 and in yarn.nodemanager.resource.memory-mb and everything worked. I was able to use example for uppertext.