home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.protocols.ibm
- Path: sparky!uunet!haven.umd.edu!decuac!pa.dec.com!engage.pko.dec.com!nntpd.lkg.dec.com!smaug.enet.dec.com!cambria
- From: cambria@smaug.enet.dec.com (Michael C. Cambria)
- Subject: Re: JES2 Performance Problem
- Message-ID: <1993Jan12.184046.22284@nntpd.lkg.dec.com>
- Sender: usenet@nntpd.lkg.dec.com (USENET News System)
- Organization: Digital Equipment Corporation
- References: <ll3v99INNl3@exodus.Eng.Sun.COM>
- Date: Tue, 12 Jan 1993 17:41:36 GMT
- Lines: 23
-
-
- In article <ll3v99INNl3@exodus.Eng.Sun.COM>, kek@bullterrier.Eng.Sun.COM (Kevin Kalajan - SunConnect Product Marketing) writes...
- >I could not post to bit.listserv.jes2-l.
- >
- >We have done exhaustive testing in trying to pinpoint why JES2
- >is waiting anywhere from 1-4 seconds between sending blocks
- >to a Sun emulating a 3770 SNA/RJE station. This includes Sniffer
- >and GTF traces that show the link idle for the above time while
- >we wait for JES2 to send the next block.
- >
- >
- >What can we do to "tune" JES2? We have gone through tuning the NCP and cannot
- >see any better results. Is there some buffer pool tuning that can be done to
- >"pre-read" disk spool data in JES2 so that spool data can be sent from the
- >buffer pool instead of waiting to read the disk? Is there internal throttle that
- >JES2 is using that can be modified?
- >
- >Any help will be appreciated.
- >
-
- Have you already ruled out outbound session pacing?
-
- /Mike
-