The following suggestions are made to assist service personnel and the customer in 3466 C00 installations when attaching to the 3494 library. The following additional parts and "script" modifications are necessary when putting the C00 and 3494 together to resemble a C30 subsystem as closely as possible: 15 m RS-232 cable, P/N 09L0764, for ADSM communications between the RS/6000 and 3494 Library Manager (quantity = 1). a. If 3590 drives are installed in both the L12 frame and D12 frame (normal for 3494s ordered separately), order: --> Two (2) 12 m SCSI cables, P/N 05H4649, to connect the two SCSI adapters in the H50 RISC to the two 3590 drives in the L12 frame of the 3494 library. Long connector screws (for hammerhead connectors) are needed for one side. --> In this configuration, two (2) 8 m SCSI cables, P/N 08L6232, should also be provided for interconnection between 3590 drives ( 2 SCSI strings) in the L12 frame and D12 frame of the 3494 library. B. If 3590 drives are installed only in the D12 frame (as in a 3466 C30 configuration), order: --> Two (2) 18 m SCSI cable, P/N 05H4650, to connect the the two SCSI adapters RS/6000 H50 to eac 3590 drive string in the D12 frame of the 3494. 3.4 m SCSI cable, P/N 05H4645, for connection between 3590 drives in a SCSI string in the D12 frame (quantity = 3). RS232/RS422 converter, P/N 50G0947, for attachment (at ARTIC ports) to the RS422 cable from the 3590 drive to ARTIC ports (quantity =3). With these cables/parts in hand, the CEs can follow the installation instructions in the Maintenance Information (MI) manual for a C30. If the cabling diagrams supplied in the MI are followed, the customer or business partner can then do the following to get the 3494 and 3590s defined to AIX and ADSM and get the tapes checked into ADSM. For the C00 these steps are NOT the CE's responsibility. run 'def3590'. This will define the 3494/3590 to AIX. This assumes the direct attach method was used. This script will fail if a cabling sequence other than that defined in the MI is used. If this fails due to cabling, then the following must be done by hand: a. /etc/methods/defatl -ctape -slibrary -tatl -l'lmcp0' -a library_name='3494a' b. mkdev -l lmcp0 c. mkdev -c tty -t 'tty' -s 'rs232' -p 'sa1' -w 's2' d. /etc/lmcpd e. cfgmgr run 'lsdev -Cc tape' to verify all the drives show up with /dev/rmt# definitions and the library manager shows up with /dev/lmcp0 definition. logon to ADSM as an administrator and do the following definitions (these are outlined in the Model C00 setup guide): a. define library 3494lib libtype=349x devi=/dev/lmcp0 b. define devclass 3590_class devtype=3590 libr=3494lib mountl=### (you must provide mount limit) c. define drive 3494lib rmt# device=/dev/rmt# (do for each drive) d. define stgpool tapepool 3590_class collocate=XXXXX reusedelay=XXXX maxscr=XXXX reclaim=100 e. define stgpool copypool 3590_class collocate=no reusedelay=XXXX maxscr=XXXX reclaim=100 f. define schedule dbbackup_full type=admin starttime=13:00:00 day=any startdate=today dur=2 duru=hours per=1 peru=days cmd='backup db devclass=3590_class type=full scratch=yes' The customer can than use the initTape script to checkin volumes into ADSM. They can also do this by hand via the 'checkin libvol' command provided by ADSM. Also very important: -- Availability of the telephone analog line to allow attachment of the modem during installation (to allow Supporttto dial in if necessary). -- At least one operating network (Gigabit Ethernet or Ethernet or Token ring, etc.)for attachment to the network adapters in the RS/6000. LABS FOR JULY UPDATES: replace.raid lab 1. How do you know this box has raid on it? 2. How can you tell if the raid arrays are in good status? 3. How does the replace raid script differe if only one disk goes bad, versus multiple disks? 4. Where is the file kept that will be used to rebuild the array? Take a look at it. 5. Go the machine and replace 2 disks in a good array with 2 NEW disks. Replace one at a time so you can answer question 5. A. How did you know what disks were in each array? B. How did you know what disks to pull? 6. What state does the array go in after one disk is pulled? After two disks are pulled? 7. Run replace.ssa 8. How do you know the script was successfull? Call_home lab 1. How can you tell if call home is installed? 2. Would this work on any box with a modem? 3. What are the three areas to be conserned about with call home, and which files are associated with each? 4. Switch the modem from call_home to call in ability. What changed? 5. What frequency will the cronttab try to change this server back? Multiple Server lab 1. Is there an entire disk array that you could move to a new server? How did you determine this? 2. If there is NOT an entire disk array, is it possible to use move.ssa to move around resources on the primrary server to free up an array? If yes, do this. 3. Determine which array, or individual resources you will dedicate to the new server. 4. Create a new server. If possible do this and select to move an entire array. 5. What size db, log and stgpool were created? Is this big enough to act as a disaster server? 6. Use move.ssa to move ONE individual resource from the primary server to the kin server. (you can move 100% or less) 7. What are the names of all the resources given to the kin server? 8. Where is this kin server located at? What is in that directory? 9. Run isitup, endadsm, startadsm so you can see how 2 servers show up. 10. Log onto the new server and see what was defined - how do you do this? 11. Once the kin server is created, remove it. 12. Where did all the resources go to on the primary server?