|
The main cause for the 'PH_DROP_EVENT_FROM_SHARED_BUFFER' error is Full Queue drops by the Workers.
The steps to follow are below:
Remove the files from the cache - To DO ON SUPER/WORKERS DURING DOWNTIME - Take a snapshot first:
cd /opt/phoenix/cache/<SUPERIP>/phoenix/rest/dataRequest/rule/
rm -f *
mv /opt/phoenix/cache/<SUPERIP> /opt/phoenix/cache/SUPERIP.old
vi /opt/phoenix/config/phoenix_config.txt -> Search for the line 'notification_server_thread_num'.
Now change this from:
notification_server_thread_num=20
To:
notification_server_thread_num=50 -> Search for 'count_distinct_precision'.
Modify the value to the below:
count_distinct_precision=9
Save the configuration and exit the file.
- Restart the following on Super:
killall -9 phMonitor phRuleWorker phRuleMaster
- Restart the following On Worker:
killall -9 phRuleWorker
- Make sure all processes are up and running:
phstatus
|
