I’m currently runing Dremio 21.x on AWS Edition, but mt coordinator runs almost on 100% CPU in some cases.
My datadog monitor above
This has crashing the entire operation and i read in the docs that AWS can’t have more than one coodinator, it’s possible to migrate from AWS Editon to YARN or other solution?
@hl-nahuz Let us first find out why your CPU is at 100%, if we are able to address this then we do not have to migrate,
- How many sources
- Types of sources
- Refresh frequency of each source
- All GC and server logs from coordinator when this issue happened
- thread dumps, script below when problem happened
- ps -ef | grep dremio
- login as the the unix id running the Dremio process
- mkdir jstack
- cd jstack
- vi jstack.sh (run jstack more frequently if the time to capture is very less)
#!/bin/bash
for i in seq -w 3 1 300
do
jstack -l dremio-pid > ThreadDump$i.txt
sleep 1
done
- chmod +x jstack.sh
./jstack.sh
tar -cvf jstack.tar *.txt
gzip jstack.tar