We recently went through a heck of a mess after our EV 6 CP3 index experienced a 2.5 hour gap in data. Luckily, we knew where it was and after tinkering for 40 days and 40 nights, we got things back on track. However, Discovery Accelerator hasn't been the same since. Small searches yield valid results, but large searches - 90, 000 items or more, cause the DA to freeze in "Accepting" mode. AND, we're seeing Windows Application Log Event 42064. Basically the DA times out. The problem is that the timeout isn't enough to kill the "Accepting" mode.
I think the index on the SQL server is corrupt, and I was thinking of running Replayindex to rebuild a new index. Any reason why I shouldn't? BTW, my logic is based on the fact that nothing has changed recently except the index. Any thoughts? Advice?