Hi Prabhakar, Thanks for your interest in Drill. Firstly, I’d ask you to please update to the latest version of Drill. 1.20.1 is three years old and we are currently on version 1.22.0. Whatever is causing this may very well have been fixed over the years as we have made numerous updates to the Hadoop libraries.
Secondly, to adequately debug this, we’d need to know what version of Hadoop you are using on both clusters, as well as your Drill configurations. We’d likely also need the full stack trace. Best, —C > On Sep 26, 2025, at 09:09, Prabhakar Bhosale <[email protected]> wrote: > > Hi Team, > > I am having drill 1.20.1 deployed on cloudera hadoop on two different > cluster. > > Cluster 1 - On This cluster the query is failing with below mentioned > different errors at different times until I increase the direct > memory setting of drill. > Cluster 2- Same query works with default memory settings for drill. > > Important to note that cluster 2 having more data as compared to cluster1 > but still querying cluster1 fails . Please note I am executing the same > query on both the nodes. > > What could be possible reasons for this behaviour? Any tips or pointers to > troubleshoot the error is highly appreciated. > > Errors > 1. Connection Error: Error setting up remote > intermediate fragment execution > 2. Resource Error: one or more nodes ran out of memory while executing the > query > 3. DATA_READ_ERROR: Error reading page data - this error repeats for > different files and for different column at different times > 4, Internal_error: Error in drill parquet reader (complex) Message : > Failure in setting up reader > > Thanks and Regards > Prabhakar
signature.asc
Description: Message signed with OpenPGP
