Well.. The storms have abated somewhat, so a summary is in order here. I managed to fix this issue.. although fix is probably not the right term. To get the sqlplus stuff to run fast in the Solaris environment I did the following. I reduced the tcp_deferred_ack_interval via ndd -set /dev/tcp tcp_deferred_ack_interval 10 This made everything sing.. Now, I'm curious about what massaging this kernel parameter will break. So far nothing.. but who knows. I'd love to get some feedback on this. Anyway, thanks to all that made suggestions.. I actually came up with this one on my own. Original Post. > We have two E250 running Solaris 7 talking to an E450 running Solaris 2.6 > and oracle 8.1.6. > We are currently trying to tune our new app that is running a new > application server.. and > we are seeing incredibly slow performance.. One avenue that we are looking > at is the > speed of the database connections from the Solaris 7 boxes to the Oracle > server. > Here's what we have.. Using an sqlplus timing script the E250's take 10 > times longer > to do the same set of select statements that an NT box can do. This is using > SQLPLUS > over the network. > > Now... I don't think our network is the issue.. Is there some gross tuning > information > that we don't have that could help us cure this situation? > > Thanks > > sueReceived on Thu Dec 13 22:51:02 2001
This archive was generated by hypermail 2.1.8 : Wed Mar 23 2016 - 16:32:37 EDT