Recently there are serious blocking issues occured in one of our product databases, though I turned ON the 1222&3605 trace flags, there are no useful information written into the error log. But I could tell from the "Blk by" field of sp_who2 that there were many sessions blocked. The number of blocked sessions were too much to find the first spid in the blocking chain. The problem disappeared without any efforts been made. My guess was: someone used one or more thread in the client application to process data, selecting a few rows from a table for example, after the blocking encountered, he or she stopped the thread secretly. The exection time for a single selection is very short, but the exection interval is too frequently, may 5~7 times per a second. If there are other sessions updating the same table, blocking occurs. Parts of my colleagues like such kind of ways to process data, which really make me have a headache. I don't know why such blockings would be recorded in the error log by SQL Server. Does anyone have a suggestion that how to find the problem session? Thanks.