Im having an ongoing issue with performance of the system, i'm on release v3.63PE (the latest release). If I use a timeseries widget and call 10,000 data points from a device, over 30 days. The system completely hangs. But if I call 10,000 data points from 10 devices (1000 from each device), the data comes back relatively quickly. The time series charts are also immediate, displaying huge amounts of data. It is odd that tables and trip animation widgets are very slow to show the data. Im unclear on the configuration options within Cassandra and TB to retrieve data as fast as possible. Essentially, the whole process of pulling data out of Cassandra seems very slow, even if it is a single telemetry field.
Pulling over 10,000 pieces of telemetry will crash the system. The system has 64GB RAM and 8 core processors, so there are plenty of resources available. All resources of the system seem under utilised but TB is struggling. Anyone resolved similar performance issues? Also are there thingsboard.conf parameters I can add, to improve performance?