It has been some time since I tracked the details. However -- I remember it being the case that just opening the HPD:Helpdesk form was 500+ queries to the DB. (One person opening the helpdesk app from the WUT)
So -- you might need some more threads :) -John On 12/14/07, Carey Matthew Black <[EMAIL PROTECTED]> wrote: > > Well.. > > In a 28 day month there are 2419200 seconds in the month. So if you > are creating 888 thousand records in that month then that would only > average 0.367 records per second. So one thread should do it if your > workload is evenly distributed across a 24x7 schedule. :) > > Now if we are more constricted and all of those numbers are only > during an 8 hour days and only 20 business days in a month then we are > talking about 576000 seconds in a month. Which would yield an average > of 1.54 records per second. So I guess two threads should deal with > that load ok. :) > > > But on a more serious note.. as others have pointed out.. what you > need will highly depend on several issues outside of issues like > "tickets created" or "processes completed" but on things like Network > delays, RDMBS performance, number of "bad" searches done by users, > etc... > > And I would suggest that there really should be no number of anythings > that should cause the server to crash. If there is such a value then > that should be bugged and fixed. :) > > I will also mention that if you do testing via "testing tools" like > "Silk" or other "LoadRunner" you may have issues due to how you do > specific things in those tools too. So be mindful of your "first > results", especially if they are really bad. I have even seen > conditions that produced a kind of "hung ARS server" when the testing > tool was not properly closing the connections to the ARS server and > was trying to connect more API clients than the ARS server had > threads. ( The problem was totally on the testing tool side. After > fixing the tool the results we arrived at were much more reasonable. > But it took about a month of poking around to try to get the "testing > tool group" to even admit that they were the problem and not the ARS > server. Just thought you should know.) > > HTH. > -- > Carey Matthew Black > Remedy Skilled Professional (RSP) > ARS = Action Request System(Remedy) > > Love, then teach > Solution = People + Process + Tools > Fast, Accurate, Cheap.... Pick two. > > > On Dec 14, 2007 2:22 PM, Tadeu Augusto Dutra Pinto <[EMAIL PROTECTED]> > wrote: > > ** > > Hi crazy friends! > > > > > > I'd like to know (if it exists) what's the maximum number of users doing > a > > action simultaneously (at the same time) in a ARS application??? > > Have a max number (of users creating records, for example) that crash > the AR > > Server? > > > > I need this information because I have to describe this in a document of > > 'Test Strategy'... and 'Requirements Specification' too... > > > > Anybody knows about this?? > > > > > > I'm using: > > S.O. - AIX 5.2 > > AR Server 7.0.01 patch 5 > > > > About my applications: > > Appication1 - create around 308.000 records/month > > > > Appication2 - create around 30.000 records/month > > > > Appication3 - create around 550 records/month > > > > > > []'s > > Att, > > Tadeu Augusto Dutra Pinto > > HSBC Brasil Outsourcing > > > _______________________________________________________________________________ > UNSUBSCRIBE or access ARSlist Archives at www.arslist.org > Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are" > -- John David Sundberg 235 East 6th Street, Suite 400B St. Paul, MN 55101 (651) 556-0930-work (651) 247-6766-cell (651) 695-8577-fax [EMAIL PROTECTED] _______________________________________________________________________________ UNSUBSCRIBE or access ARSlist Archives at www.arslist.org Platinum Sponsor: www.rmsportal.com ARSlist: "Where the Answers Are"
