[RBASE-L] - RE: oterro processing vs r:base

[Alan Wolfe]

might not be true anymore (we're on oterro 2.6) but i heard that in r:tango r:base treats the connection to witango as if it's a single user doing a whole lot of work.  If this is the case, doesn't it only do one statement at a time?   I tested this so it apears this is what's happening but might be something else giving the same symptoms.  What i did was fire off something that took 240 seconds to execute from witango (the sql was what took the bulk of the time) and then tried to log in as a second user and do something else.  It wouldn't let me log in until the 240 second report was done (i had it send an email when it was done so i could test that).   ----- Original Message ----- From: Troy Sosamon To: RBASE-L Mailing List Sent: Tuesday, February 15, 2005 8:15 PM Subject: [RBASE-L] - RE: oterro processing vs r:base Alan,   I have never had this problem. I have specifically done multi user performance tests, not had this problem until you start to overload the server resources.   You should make sure you have the latest version of the Oterro driver.   When you do this from R:base, you are running in a different environment. When R:tango runs, you have a Single Oterro thread between the Tango Server and the R:base database, no matter how many people hit it.  It will process multiple requests at once and return the results as they complete.  This depends on server resources mainly memory.   In R:base, you have several different machines running programs against a single db and unless you have concurrency locking issues, they all run at once.  R:base does the processing from the workstation and not at the server.  It just needs the disk access resources from the server.   Troy     From: rbase-l@rbase.com [mailto:rbase-l@rbase.com] On Behalf Of Alan Wolfe Sent: Tuesday, February 15, 2005 1:15 PM To: RBASE-L Mailing List Subject: [RBASE-L] - oterro processing vs r:base   Hi everyone,   We are using r:tango and it looks like it can only process one database request at a time.  What i mean is like if we run a report that takes a couple minutes to generate, it wont let anyone else do anything that requires a connection to that database until that request is finished.   When you do this kind of thing straight through r:base (like by running "rbg65 blah.cmd") is it able to handle more than 1 request at a time?  We were thinking about doing batch reports through r:base to overcome the problem of having a large report dominate the database for a few minutes, does anyone know if this would work or have any other ideas to get around this problem?   Alan