$Id: README.TXT,v 1.7 2001/06/08 23:36:52 mtran Exp $ Some features are tested on basic cases, however nothing has tested formally.! Sony shall not be liable for errors contained herein, or indirect, special, incidental or consequential damages in connection with the furnishing, performance, or use of this document! Incaback is a backup ultility for the Sony AIT Auto-Loader. Project Homepage: http://mookie.asd.sel.sony.com/ina/inca/ Prerequisites: -------------- 1. mtx-1.2.11 (download from ftp://ftp.estinc.com/pub/unsupported/mtx-1.2.11.tar.gz) 2. sg_utils-0.94 (download from http:/gear.torque.net/sg/p/sg_utils-0.94.tgz) 3. tar (GNU tar) 1.13.17 or newer 4. mt 5. inca-1.xx.tar.gz 6. micinfo-1.xx.tar Installation: ------------- 1. Obtain and unzip/untar the sg_utils-0.94.tgz distribution. Go to sg_utils-0.94 folder. Type 'make' then 'make install'. This will install the sg_map binary file to usr/local/bin. 2. Obtain and unzip/untar the mtx-1.2.11.tar.gz distribution. 3. In order to compile and use the micinfo utility, make sure mtx (version 1.2.11) is downloaded. Obtain the Sony distribution file, micinfo-1.xx.tar. This file can be downloaded from the project homepage. - Place micinfo-1.xx.tar in the mtx-1.2.11 source directory. - Change to the above directory and issue the following: tar -xvf micinfo-1.xx.tar - The above step will extract several files, including micinstall.sh. Run micinstall.sh to apply the patch and compile the executables. 4. Obtain and unzip/untar inca-1.xx.tar.gz into the desired directory. This file can be downloaded from the project homepage. 5. Go to the inca directory and run Install.sh script. You will be prompted for the IncaBack installation directory. Setup: ------ Before you can start backing up your system, the IncaBack setup script must be executed. This file may be found in the IncaBack installation directory (see step 5 in the Installation section above). Backup: ------- The backup script (backup.sh) can be run via two methods. 1) ./backup.sh 2) ./backup.sh with command-line options Both methods require that an .inca_pref file be present, backup.sh will not run if this file cannot be found. This file is created by running Setup.sh and contains preferences and parameters that are needed by backup.sh. This file will be placed in the INCA directory that you specify when you run Setup.sh (default is /opt/inca). The location of the INCA directory is stored as an environmental in /etc/profile (INCADIR). The first method relies completely on the values contained in .inca_pref. This is useful for scheduled backup jobs or for non-power users who only want to setup the backup parameters once. The second method allows command-line options. The options are: -f non-rewind device file, i.e. /dev/nst0 -d directories or files, i.e. /home (if multiple directories or files need to backed up then they must be enclosed by quotes, i.e. -d "/home /usr /sbin" -m backup mode, valid options are either F(ull) or I(ncremental). The first backup on a system must be a Full, subsequent backups may be Incremental. -n date in a GNU Tar recognized format. This date is used for incremental backups. -h commnad-line help Any, all, or no command-line options may be used, but any values not passed via command-line arguements will be taken from the .inca_pref file. So incomplete command-line arguments may cause backup.sh to not behave as you expect. The .inca_pref file may be manually edited after its creation for tweaking. See .inca_pref for comments. Scheduling Scheduling is handled by the cron service. Setup.sh will create a script in /etc/cron.daily to run backups daily. Immediate backups can be done bye using backup.sh from the installed directory. Logging The INCA backup package creates log files in the default location /var/log/inca or whatever was entered by the user when Setup.sh was run. The following log files are generated: setinfo.log ----------- This is a rudimentary catalog that contains backup set info in the following format: 989340721,05/08/01,09:52:01,F,1000036452,Complete 989340781,05/08/01,09:53:01,I,1000036452,Complete The comma delimited fields are: jobID ,date ,time ,mode, tape SN, status -jobID is a unique number generated by backup.sh when the job is run based on epoch time. -date is the date when the job was started. -time is the time when the job was started. -mode is whether the job was a F(ull) or I(ncremental) backup. -tape SN is the mic serial number of the tape that the job was written to. -status field contains one of three things: Complete, Error, or it may be empty which means the job may not have completed yet or it is hung. timeinfo.log ------------ This is just a log of date and time when the jobs were run. This could be retrieved from setinfo.log as it is redundant. incadump.log ------------ incadump.log contains the values of most of the variables used by backup.sh. This always contains the values from the latest job executed. It's used for debugging. job log files ------------- Each job that is run will also create a log file in the format jobID.log so from the previous setinfo.log information we can see that /var/log/inca should also contain two job log files called 989340721.log and 989340781.log. The format of these files are: 989341500 Volume 1 home/nate/inca/ home/nate/inca/CVS/ home/nate/inca/CVS/Root home/nate/inca/CVS/Repository home/nate/inca/CVS/Entries home/nate/inca/doc/ . . . home/nate/inca/doc/gnu/ home/nate/inca/doc/gnu/CVS/ home/nate/inca/doc/gnu/CVS/Root This is basically a tree of the tar verbose output. The first line contains the tar archive header which we have filled with the unique jobID. Volume 1 tells us that it may be part of a multi-volume tar archive. The rest of the file contains a listing of the directories and files that were backed up.