Before only sequential kind of workload was supporting partition offset, the uniform and timeseries were not.
So, make those other 2 workload types start supporting partition offsets for being able to write different data ranges
from different scylla-bench instances.
Caution:
using 'timeseries' workload, number of partitions may differ
from what is specified in the '-partition-count=x' parameter.
So, make sure that partition offset is big emough to avoid index
overlapping.
Before only
sequentialkind of workload was supporting partition offset, theuniformandtimeserieswere not.So, make those other 2 workload types start supporting partition offsets for being able to write different data ranges from different scylla-bench instances.
Caution: using 'timeseries' workload, number of partitions may differ from what is specified in the '-partition-count=x' parameter. So, make sure that partition offset is big emough to avoid index overlapping.