hey just bumping this message in case anyone has a...
# troubleshootingl
Luis Fernandez
12/17/2021, 7:15 PMhey just bumping this message in case anyone has any idea as to what’s happening https://apache-pinot.slack.com/archives/C011C9JHN7R/p1639691077070700
m
Mayank
12/17/2021, 7:29 PM
The intermittent spikes are typically because when creating and committing segments the ingestion is temporarily on hold. Once segments are committed, the consumer tries to catch up.
Mayank
12/17/2021, 7:30 PM
I haven't seen the ingestion spike causing the issue, but there could be a few things you may want to check:
Copy code
1. Any errors during segment creation (any OOMs)?
2. Query workload during the time (also CPU/Mem load).l
Luis Fernandez
12/17/2021, 7:34 PMsorry it seems like the ingestion has calmed down so I don’t think that’s an issue anymore at least for now
Luis Fernandez
12/17/2021, 7:34 PMLuis Fernandez
12/17/2021, 7:34 PMhowever for some queries i’m getting that arrayoutofbounds exception
Luis Fernandez
12/17/2021, 7:37 PMthe bigger the span of time the more i get it
m
Mayank
12/18/2021, 2:35 PM
The arrayoutorbound seems troubling cc: @Jackie. Separately, Is your time column in millis? Do you care about querying millis granularity? If no, then may be use seconds/hours/days (lowest you will query)?
l
Luis Fernandez
12/18/2021, 2:36 PMit is in seconds
Luis Fernandez
12/18/2021, 2:40 PMand this is the lowest I will query
👍 1