Re: Performance penalty of multiple UPDATEs of non-pk columns

2015-05-22 Thread Artur Siekielski
Thanks, I wasn't sure if memtables and sstables contain only the newest values (I though replication might require storing old values). So the number of lookups for a newest value should be bound by max_compaction_threshold setting. Looks to me it's safe to perform many UPDATEs of non-pk colum

Re: Performance penalty of multiple UPDATEs of non-pk columns

2015-05-21 Thread Sebastian Estevez
Counters differ significantly between 2.0 and 2.1 ( https://issues.apache.org/jira/browse/CASSANDRA-6405 among others). But in both scenarios, you will pay more for counter reconciles and compactions vs. regular updates. The final counter performance fix will come with CASSANDRA-6506. For details

Re: Performance penalty of multiple UPDATEs of non-pk columns

2015-05-21 Thread Jens Rantil
Artur, That's not entirely true. Writes to Cassandra are first written to a memtable (in-memory table) which is periodically flushed to disk. If multiple writes are coming in before the flush, then only a single record will be written to the disk/sstable. If your have writes that aren't coming wit

Performance penalty of multiple UPDATEs of non-pk columns

2015-05-15 Thread Artur Siekielski
I've seen some discussions about the topic on the list recently, but I would like to get more clear answers. Given the table: CREATE TABLE t1 ( f1 text, f2 text, f3 text, PRIMARY KEY(f1, f2) ); and assuming I will execute UPDATE of f3 multiple times (say, 1000)