home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
OS/2 Shareware BBS: 18 REXX
/
18-REXX.zip
/
netbck01.zip
/
READ.ME
< prev
Wrap
Text File
|
1994-05-20
|
8KB
|
185 lines
This package is an OS/2 v2.x REXX based script to support Lan server
PEER backups from a server using Sytos for OS/2.
------------------------------------------------------------------
NOTE: I make no claims that this code will work for you as is. Its in use
in three of our LAN sites and works. Feel free to modify
the exec in eny way you want.. You're on your own.
I'm no LAN guru so be gentle if you find glaring 'silly stuff' in
this code..
For those who want, I can be reached at:
Compu$erve: 73677,1167
Internet: Larry.Kalyniak@WPCUsrGrp.MB.CA
---------------------------------------------------------------------
Files included should be:
NETBACK DAT 89 5-18-94 2:32p
NETBACK CMD 28718 5-18-94 2:28p
NEWTAPE MAC 291 4-11-94 9:38a
APPEND MAC 294 4-11-94 9:39a
READ ME
---------------------------------------------------------------------
Required for operation:
- SYTOS for OS/2
- IBM LAN Server 2.x/3.x
- PEER services installed on the required workstations.
---------------------------------------------------------------------
Installation:
1) Copy the NETBACK.CMD file into the SYPLUS directory.
2) Copy the NEWTAPE.MAC and APPEND.MAC Sytos macros into
the SYPLUS\MAC subdirectory.
3) Edit the NETBACK.CMD REXX file and customize the 4 lines
(delimited by comment blocks at the start of the file)
to specify your disk drive, Sytos directory,admin ID
and admin password used for your PEER backups.
4) Create a backup schedule file (in the same directory)
indicating users to be backed up. (file: NETBACK.DAT)
5) Add the NETBACK cmd file to some form of execution scheduler
which allows VIO windowed apps.
---------------------------------------------------------------------
Notes
---------------------------------------------------------------------
Theory of operation:
NETBACK.CMD builds a list of scheduled users to backup
based on the NETBACK.DAT file. The network is then queried
if these users are currently available (workstation up and
running peer services.). All available workstations are then
queried for defined PEER disks. These disks are then
added to a final PENDING backup list. During this time
users are sent warning messages indicating that a backup is about
to take place.
At this point NETBACK accesses the FIRST user's first disk
waiting for backup and links to them under the define 'RemoteDiskId'.
A copy of the NEWTAPE.MAC is made (named CURRENT.MAC). Into this
copy is written the user ID and DISK designation for this
backup set. The disk is then backed up using the UNATTENDED
backup mode of SYTOS using the CURRENT.MAC.
The same procedure is followed for all subsequent pending
user/disk backup combinations. The only difference is that
after the first backup onto the tape, NETBACK uses the APPEND.MAC
which is defined to APPEND backup sets to the current tape.
As each user is successfully backed up, NETBACK then sends
a net msg to each indicating when it is safe to shutdown.
On completion, NETBACK logs off the specified ADMIN ID.
---------------------------------------------------------------------
Comments:
During operation, NETBACK.CMD maintains a log/history
of who was backed up, what drives were backed up and the
START/END times of each backup.
All LAN cmds are called via cmdline NET.EXE with the output
spooled to, and read from, a QUEUE. It is NOT a private
QUEUE so if you have other REXX apps running using QUEUES, you may have
to change this to prevent data corruption.
---------------------------------------------------------------------
KNOWN PROBLEMS:
* On occation SYTOS will be in the middle of a backup and it
will just freeze, waiting..(forever) It 'appears' as if the connection
to the PEER drive has been disconnected for some reason. SYTOS never
receives the error and just sits waiting for the drive data.
I still haven't figured out exactly what the problem is because it
occurs at random. (At least appears to.) Note that our server(s) are
also running LTLW, DBM/RDS, and as a NN for APPC/CPI-C apps so its
possible that we're being effected from 'outside' the Lan server/SYTOS
system.
* For whatever reason, I can't get reliable backups if I define
the NEWTAPE.MAC/APPEND.MAC to use SYTOS's QFA (Quick File Access)
during the backup.
I'm guilty of having read in some FIDO notes that other folks had
problems with QFA, so I haven't tried very hard once it failed for me
the first few times.
* I can't/don't check for TAPE FULL conditions simply because it has
never been an issue. (Our LANs are have 10-15 workstations and a 4G DAT
tape on the server. There also isn't anyone around at night to change
tapes if required.<g>)
If you need/add this feature, drop me a line.. The Schedule/Group ID
does allow you to 'stagger' the backups. (Backup different groups on
different nights)
---------------------------------------------------------------------
Format of the NETBACK.DAT schedule file:
* One line per user
* 1st word: (blank delimited): Schedule group id.
2nd word: User LOGON ID
3rd word: Requester ID/Workstation ID
4th word: Nickname used for messages
---------------------------------------------------------------------
ie:
1 LSMITH VAN1R00E Lenny
9 TJONES VAN1R003 Tommy
---------------------------------------------------------------------
Schedule Netback accepts a single cmdline parameter.
group ID: This is a NUMBER indicating which group within
the NETBACK.DAT schedule should now be backed up.
When NETBACK reads in the NETBACK.DAT backup
schedule, it does one of two things:
NO CMDLINE GroupNumber passed: Use entire
backup list.
CMDLINE GroupNumber passed: Only backup users with
a matching group number.
---------------------------------------------------------------------
LAN Server AT.EXE: LAN Server has a scheduler called
AT.EXE. This package *WILL NOT* work with this
scheduler. AT runs fully disconnected.. No
access to SCREEN, KEYBOARD resources. => SYTOS
will not run under it..
---------------------------------------------------------------------
Installing PEER services on requesters: See your LAN manuals. Just
make sure you give Read,Write,Create, Attbribute and
Permissions rights to the ID you define on the requester
which under which the backup system will gain access
to your drives.
(Don't forget to SHARE the drives once PEER is
installed on the workstation.. Add that to the
user's STARTUP.CMD..)
-----------------------------------------------------------------------
Hints:
* I tend to let the lan admin workstation be the LAST entry
in the NETBACK.DAT backup schedule. When this person
gets in, he/she can immed. tell if the Lan backup was successfull
simply by checking if he/she received the SUCCESSFULL
backup msg for their own workstation.
* Checking the time difference between the STARTING BACKUP msg
and the COMPLETED BACKUP msg can give an indication as to
the success of the backup.
<EOF>