What is causing log file sync and gc current block 2-way waits
I have a two node RAC cluster with a very low latency OLTP application running on it. We have average application transactions of about 100 ms (we have instrumented our code to track application transaction times). However, periodically we will see transaction times spike to between 20,000-50,000 ms (20-50 sec)! When this happens it seems like transactions start piling up and then all of a sudden they will clear. AWR reports and ASH analytics from Enterprise Manager show there are "log file sync" waits and "gc current block 2-way" waits. Today, this happened, but it only occurred with one application