From owner-freebsd-questions Fri Mar 3 7:44:48 2000 Delivered-To: freebsd-questions@freebsd.org Received: from yoda.vardus.net (yoda.vardus.net [195.153.248.50]) by hub.freebsd.org (Postfix) with SMTP id 1960137B5C8 for ; Fri, 3 Mar 2000 07:44:44 -0800 (PST) (envelope-from alissa@vardus.net) Received: (qmail 12310 invoked by uid 1009); 3 Mar 2000 15:45:05 -0000 Date: Fri, 3 Mar 2000 15:45:05 +0000 From: Alissa To: freebsd-questions@freebsd.org Subject: Smart UPS 3000 and ShareUPS Message-ID: <20000303154505.A12272@yoda.vardus.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.95.4i Sender: owner-freebsd-questions@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG All - Sorry about the length of this, but I'm getting desperate. I am trying to connect 2 FreeBSD machines to a Smart UPS 3000 ups via a ShareUPS, one is 3.4-RELEASE, the other is 2.2.8-RELEASE. At the moment, I would be happy just to get my test server (3.4-RELEASE) connected. I have tried upsd, upsmon, nut, and Powercute Plus (for both SCO and Linux), with no success. I have connected the black cable (can't remember the part number at the moment) from the machine's serial port (COM1) to a basic control port on the ShareUPS, as we have an NT machine on the Advanced (UPS control) port that's doing it's job reasonably well. Just to be safe, I tested the cable I am using on the FreeBSD machine on the NT box, and it saw the UPS just fine. As the information I have managed to find in newsgroups, mailing lists, etc was unclear about what device to use, I have tried everything using both ttyd0 and cuaa0 (separately, of course) upsmon: The daemon starts up, but upsstat gives null values and causes upsmond to core dump (signal 11) for both /dev/ttyd0 and /dev/cuaa0 upsd: daemon starts up, but shows null values in it's status report to syslog, and says it cannot put UPS into smart mode for both /dev/ttyd0 and cuaa0 nut: for both /dev/ttyd0 and /dev/cuaa0, apcsmart reports ... su-2.03# apcsmart /dev/cuaa0 Network UPS Tools - APC Smart protocol driver 0.42 (0.43.0) Unable to detect an APC Smart protocol UPS on port /dev/cuaa0 Check the cabling, port name or model name and try again /dev/ttyd0 reports Permission denied. changing the owner of ttyd0 to nobody (the running uid) and rerunning apcsmart /dev/ttyd0 yields: su-2.03# chown nobody /dev/ttyd0 su-2.03# /usr/local/sbin/apcsmart /dev/ttyd0 Network UPS Tools - APC Smart protocol driver 0.42 (0.43.0) Unable to detect an APC Smart protocol UPS on port /dev/ttyd0 Check the cabling, port name or model name and try again Powerchute: I tried both Linux and SCO binaries. Both installed fine, but could not communicate with either /dev/ttyd0 or /dev/cuaa0. With cuaa0, this error was reported when the install script was testing for the local control port: WARNING: Port must be configured for local control in order to work with a "Smart" Signaling UPS With ttyd0, this error did not occur and it was reported correctly as the local control port. However, when it came time for the install script to check serial communications, it reported the same thing for both devices: ---------------------------------------------------------------------------- WARNING: Could not communicate with device on /dev/ttyd0. Check that the communications cable is attached properly to both the device and the serial port. \c ---------------------------------------------------------------------------- NOTE: Though the PowerChute Plus for Unix installation is complete, if you quit before verifying communications, the application may not work properly ---------------------------------------------------------------------------- Do you wish to Exit or Retry communications? [R/E] \c As far as communicating with the UPS through the serial port, using cu: su-2.03# cu -s 2400 -e -o -h -l /dev/ttyd0 Connected. However, it does not appear to be truly connected, as I can type whatever I like and it shows up on screen, but I am unable to break out of it and the process must be killed from another session. I get the exact same behavior with /dev/cuaa0. (I have so far been unable to find what it should be doing, if anything) Checking the ps listing shows 2 identical cu processes running, with consecutive pids. kill -9 on the first one brings me back to shell on the locked terminal, but the 2nd process must be manually killed as well, or it just keeps running. This happens on every machine I try it on, btw. As far as getty goes, I have tried removing the ttyd* lines from /etc/ttys (then reinstalling every program one at a time, uninstalling each after it bombed, and rebooting after each uninstall), I have also tried a variety of different options for the entry (with the same uninstall/reboot/install procedure going). Unfortunately, I have very little experience when it comes to serial links, so it's a bit like stabbing in the dark. If anyone can help me out, t'would be much appreciated :) Cheers, Alissa Rogers To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-questions" in the body of the message