search

GoogleCloudPlatform / PerfKitBenchmarker

PerfKit Benchmarker (PKB) contains a set of benchmarks to measure and compare cloud offerings. The benchmarks use default settings to reflect what most users will see. PerfKit Benchmarker is licensed under the Apache 2 license terms. Please make sure to read, understand and agree to the terms of the LICENSE and CONTRIBUTING files before proceeding.
https://googlecloudplatform.github.io/PerfKitBenchmarker/
Apache License 2.0
1.91k stars 512 forks source link

Hadoop in Azure stuck in ACCEPTED state #744

Open jldiaz opened 8 years ago

jldiaz commented 8 years ago

I ran

./pkb.py --cloud=Azure --benchmarks=hadoop_terasort

When the script finally reached the point in which it invokes yarn to run TeraGen, it got stuck for hours wating it to end.

I logged via ssh in the master machine and used yarn to discover that, the TeraGen application was ACCEPTED but still not RUNNING after 11hours.

No idea about what the issue could be. Insufficient memory? Which instance type is required as minimum to run this benchmark? By default pkb uses Small, which translates into A1-Standard (1 Core, 1.75Gb RAM)

ehankland commented 8 years ago

Thanks for reporting this. The current configuration requires at least 3GB of memory per machine (see https://github.com/GoogleCloudPlatform/PerfKitBenchmarker/blob/master/perfkitbenchmarker/data/hadoop/mapred-site.xml#L27). I think this was likely copied from the setup instructions here: https://hadoop.apache.org/docs/current/hadoop-project-dist/hadoop-common/ClusterSetup.html

It shouldn't be hard to make this run (or at the very least warn the user if they attempt to run it on a machine with less memory). Until this is fixed, running on any machine type with more than 3GB of RAM (e.g. the Standard_D1 type) should work.