Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Posting-Version: version B 2.10.3 4.3bsd-beta 6/6/85; site ucbvax.BERKELEY.EDU Path: utzoo!decvax!ucbvax!info-vax From: linscomb@NGP.UTEXAS.EDU (Thomas J. Linscomb) Newsgroups: mod.computers.vax Subject: Re: "Network partner exited" message Message-ID: <8601040707.AA27274@ngp.UTEXAS.EDU> Date: Sat, 4-Jan-86 02:07:11 EST Article-I.D.: ngp.8601040707.AA27274 Posted: Sat Jan 4 02:07:11 1986 Date-Received: Sat, 4-Jan-86 15:40:54 EST Sender: daemon@ucbvax.BERKELEY.EDU Organization: The ARPA Internet Lines: 51 Approved: info-vax@sri-kl.arpa > > We are a VAX site running version 4.1 of VMS. We have jnet version 1.2 > RSCS emulation software to enable our link to BITNET. From time to time, > users sending network messages from within the VMS Mail Utility are confronted > with the following message: > %MAIL-E-LOGLINK, error creating network link to node 0 > -System-F-Linkexit, network partner exited > After re-trying the same message 2 or 3 times the error goes away and the > message is successfully sent, but this is annoying. > The System Messages and Recovery Procedures Manual says this message means > the log file on the remote node was not open or was write locked. Does this > make sense? Does jnet really send a signal to the target node to check if > the "log file" is available, before it lets the user send a message to that > node? Has anyone else seen this message? > > Regards, > Robert Venard > Hi Robert, Under VMS v3.x, v4.0 and v4.1 we (and other users on info-vax) also had a similiar problem with DECnet. In our case it was not with JNET but your problem sounds real familiar. Below is the explanation I received from Digital. I have rerun our failing command files and it went from a 1 in 30 failure rate to no failures in over 6000 executions under VMS v4.2. Thank you Digital. As to your specific JNET questions I can only say that JNET does in some cases open DECnet links to your own system (node 0) to startup some things. I remember sending mail to a BITNET address as one of these cases. This is how you are getting DECnet error messages while sending mail messages. Since I am not familiar with JNET I can't say much more then that. --Thomas -------------------- Answer for QAR #00460: A problem was found which caused a node receiving a [DECnet] connect request to occasionally (and incorrectly) ignore it as a duplicate. This would occur if the link ID was the same as an existing link from a different node with a higher address. The result was most visible with applications which opened large numbers of links one after the other, since whenever they got around to the same link ID, they would time out. The problem is exacerbated by a second bug which caused the sequence number field of the link ID to not be incremented. Fixed in V4.2. -------------------- PS: Please say hello to Candy and Scott for me. --Thomas Linscomb aka linscomb@ngp.UTEXAS.EDU aka cctlinscomb@agl.UTEXAS.EDU --Advanced Graphics Lab/Computation Center/The University of Texas --Austin, TX 78712 phone: 512/471-3241 --uucp: ...seismo!ut-sally!ut-ngp!linscomb linscomb@ut-ngp.UUCP --bitnet: cctj001@utadnx