I have a basically inactive (so far) 4 node cassandra cluster that connects to opscenter. There is a table that writes very little in the last week or so (test cluster). It works 2.1.0. It happened with ssh in and out of curiosity, du -sh * was launched in the data directory. Here is what I get:
4.2G commitlog 851M data 188K saved_caches
The commit log directory contains 136 files. I blushed and then drained the cassandra, stopped and began the service. These files still exist. What is the best way to get rid of them? Most of the material is related to opscenter, and I tend to just blow them away because I don't need test data. Wondering what to do if it pops up again. Rate all the tips.
ashic source share