First, thanks to those who wrote to me with suggestions and thoughts, including Julie Peers David Eisner Wanke Matthias Marco Breedeveld Joe Fletcher Luc Suryo I have to say that this issue did not get resolved the way I would have liked. What I ended up doing, as an experiment, was to connect the Neo library to a Linux server obtained from Western Scientific that was host to an external LVD SCSI RAID enclosure. With the terminator on the jukebox now lit green, reporting an LVD connection, Networker was able to find and communicate with the jukebox enough to install it. What I am forced to assume is that the SE-SCSI port on the E6500 created such a poor connection to the jukebox (whose end was a VHDCI Micro Centronics 68-pin) and its drives that, while the devices were all detected on boot and the tape drives could be read from/written to, Networker could not communicate well enough with the jukebox to do anything with it other than see it. We are going to get one of the two jukeboxes into production over SCSI with a Linux node fitted with an Adaptec 29160. We are going to try to get the optional jukebox fiber connection working with the 6500 using the other library, since the SCSI connection is a no-go with that server and we don't want to invest in another expensive sbus card. - Eric ------- Original Post: This is something of a shot in the dark, but we're caught in a bad situation and my hope is someone reading this may have worked his/her way out of a similar situation. We are trying to get an Overland Neo LXN4000 series jukebox with two HP LTO drives working with an E6500 server. The drives and jukebox are connected to one another using VHDCI micro SCSI connections, and the box is then connected to the server using a wide 68-pin-to-68-pin-VHDCI cable. The terminator on the end of the chain on the jukebox is lit red indicating an SE connection. The server, using tar, can easily write and read to the drives. The MTX program (mtx.sourceforge.net) can communicate with the jukebox and move tapes around, including to and from the drives. The problem starts with Legato Networker. The software cannot seem to be able to communicate with the jukebox robotics. jbconfig finds the jukebox but after confirming the SCSI control port it simply says it's unable to access the jukebox. The sjirdtag program spits out bash-2.04# /etc/LGTOuscsi/sjirdtag scsidev@7.6.0 SJIRDP: Device busy Code:0x29, Str=<I/O error> Legato support has been communicating with me (slowly) for the last couple of weeks, but we've gotten nowhere. I'm using version 6.1.2. I've been hoping they could tell me about a good debugging program to try but nothing has given any useful information. Overland has been helpful but they believe it is a Legato issue, especially since tar and mtx have no issues with the system. The only reason I doubt that a little is when I reset the server and drop to the ok prompt, probe-scsi-all reports the HP drives properly, but when it is supposed to report the Overland (OVERLANDNEO Series), it says Unknown message 2 I am open to any and all theories, suggestions, and similar war stories on this kind of thing. I'm just looking for new ideas at this point. _______________________________________________ sunmanagers mailing list sunmanagers@sunmanagers.org http://www.sunmanagers.org/mailman/listinfo/sunmanagersReceived on Mon Dec 9 16:20:58 2002
This archive was generated by hypermail 2.1.8 : Thu Mar 03 2016 - 06:43:00 EST