Re: Shared buffers, db transactions commited, and write IO on Solaris

Erik,

You'er welcome!  However, I believe our situation is very different

from what you're testing if I understand you correctly.  Are you

saying that you're entire database will fit in memory?  If so, then

these are very different situations as there is no way ours could

ever do that.  In fact, I'm not sure that forcedirectio would really

net you any gain in that situation as the IO service time will be

basically nil if the filesystem cache doesn't have to page which I

would think is why your seeing what you are.

Even more interesting.  I guess we've been doing too much work with 

benchmark workloads, which tend to be smaller databases.  

Thing is, there's *always* I/O for a read/write database.  If nothing else, 

updates have to be synched to disk.

Anyway ... regarding the mystery transactions ... are you certain that it's 

not your application?  I can imagine that, if your app has a fairly tight 

retry interval for database non-response, that I/O sluggishness could 

result in commit attempts spinning out of control.