Troubleshooting Big Brother

Click your symptom in the list below:

I get the message: bb: CAN'T CONNECT TO bbd

This indicates that an instance of the bb executable can't connect to the Big Brother daemon. This might be because the bbd daemon isn't running, or because the bb program cannot determine where bbd lives because of trouble with the etc/bb-hosts file.

Check the following things:

You can check for some errors in bb-hosts by running these commands:

cd etc
./bbchkcfg.sh
./bbchkhosts.sh

 

I'm running AIX: I get lots of processes, then Big Brother dies

Make sure the install_assist program is either commented out or removed from /etc/inittab. You may have to reboot your server.

I get garbage / my environment all over my screen

This is almost always due to a problem in the etc/bb-hosts file.  Big Brother needs this file to be perfect to work, and any little problem with it will cause Big Brother to fail. The most common cause of this problem is pop3 being defined as pop-3 in /etc/services.  Make sure the spelling of all services in bb-hosts matches /etc/services. Also make sure that any BBDISPLAY or BBPAGER is defined only once in bb-hosts.  Also make sure that the hostnames defined in bb-hosts are the same as returned by uname -n.

I get the message: Can't open stream socket

This means bbd cannot attach itself to port 1984 and begin listening.  Make sure there are no bb or bbd processes running; if there are, kill them. Also, make sure port 1984 is not in use by another process. Issue this command:

netstat -an | grep 1984

If anything comes back, wait a few minutes and try again. Once this command returns nothing, you should be able to start Big Brother.

The conn (connection) test is not working or always returns red

The connections column is generated from the machine defined as BBNET in bb-hosts.  This machine tries to ping every IP address listed in the bb-hosts file. Big Brother looks to see that the reply from ping contains the string bytes from. Check that the PING and PINGPARS parameters are set correctly in etc/bbsys.sh or etc/bbsys.local. PING must be the path to the ping command; PINGPAR1 and PINGPAR2 are parameters to pass to the ping command before and after the IP address. They vary depending on the operating system.

You may also want to adjust the number of packets sent for the ping test by changing the PINGTESTS parameter in etc/bbdef-server.sh on the BBNET server. Change this to 2 and analyze the results. If this does not make a difference, try 3. Note that sending more packets adds a one-second sleep between each packet. This will slow down your whole network test by [(n-1) * hosts] seconds.  If you test a large number of hosts, you will see a noticeable difference in how often your hosts get tested.

The http test is not working or always returns red

This is usually because the http test is not for the same machine as defined on that line in the etc/bb-hosts file. For example:

Wrong: 204.101.110.101 fred.bobo.com # http://youre.bobo.com/

Right: 204.101.110.101 fred.bobo.com # http://fred.bobo.com/

 

The background color is always red / yellow / wrong

The background color should reflect the most serious state on your network at any given time.  If it is not doing this, or the background color is wrong, it is because there are some leftover log files in the $BBLOGS directory (BBLOGS is defined in etc/bbinc.sh).  To check this, click the Condensed View button on the main Big Brother Web page, and the offending entries should become visible. Use the bin/bbrm command to delete them from $BBLOGS; see Removing Old Data. There may also be an HTML version in www/html.

Problems with pager notifications

The paging subsystem is very time-sensitive.  It is possible that the timing is either too long or too short for your pager.

The etc/numeric.scr file contains the commands that actually dial the pager number. The command is something like this:

dial /@[3],,,,,,,,/@[4]

The commas are used by modems for short delays during dialing. You may need to add or remove commas to adjust the timing. If you're using an 800 number paging service, you may have to embed commas in the pager number itself, as in this example:

PAGER="1800PAGENET,,,,,,7777"

Where 7777 is your account number.

The screen is purple after I've removed a host from etc/bb-hosts

After you remove a host from etc/bb-hosts, you must remove the log files for it from $BBLOGS, www/html and $BBHIST. You can do this using the bin/bbrm command.  See Removing Old Data.

The gif files for the colored dots aren't displayed properly

Make sure the BBWEB value is correctly defined in bbdef.sh as BBWEB="/bb" where /bb is the location in your web server docs directory. Check the spelling and location and correct it if needed

Availability reporting is not working

If you get a Page not found error when running the availability reporting feature, you probably have an invalid group name set for $BBHOME/www/rep. The group name of $BBHOME/www/rep must be set to the group ID of the user that the Web server is running as. Do not set the $BBHOME/www/rep permissions to 777 as this may represent a security risk. Only set the group name of $BBHOME/www/rep.

I get the message Forbidden - You don't have permission to do ... on this server

The message is coming from your Web server, and is probably correct. There are three distinct times when this problem can occur.

See the correct section below for your symptom.

Cannot view the main display

If you cannot view a Big Brother status screen after installation, first make sure that Big Brother is working and has actually generated something for you to look at. Look in $BBHOME/www for the files bb.html and bb2.html. These should exist and have a size greater than 0. If these files do not exist, your problem lies elsewhere. Follow the troubleshooting steps in the Big Brother UNIX Server  Getting Started Guide.

If the files do exist, but your Web server is not letting you see them, you most likely need to do one of the following:

Problems with reports

If you see the error message bb-histlog.sh lacks permissions to access Report directory when trying to run an availability report, it is because the Web server cannot write to the $BBHOME/www/rep directory.

The actual report output is a dynamically-created Web page that lives in that directory. Therefore the user ID your Web server is running as needs permission to write to that directory. Entering this command should fix the problem:

chown nobody $BBHOME/www/rep

History or other CGI permissions problem

First, check your Web server's error log. If you find a relevant error message, fixing the problem is usually straightforward.

Second, check the permissions on the file you're trying to execute. You need to have execute permissions on the files in your cgi- directory.

Finally, try running the command under the user ID of your Web server. If this does not work check the permissions for the user ID your Web server is using (frequently nobody), for all the directories between the root directory (/) and the cgi- directory. If any of these directories don't allow access to the Web server, the Web server won't be able to execute anything.

A simple way to test this is to su to the user ID of your Web server. Try to change directories to your cgi-directory. If you cannot, you have found the problem and will have to adjust permissions on one or more directories.

If you can cd to your cgi-directory, then try to manually execute the command. It should fail, just like it did for the Web server. Check the permissions on every directory in the path to the cgi- directory, and on the file itself. The Web server user ID must have at least read/execute permission for each directory and the file. You may have log in as root and issue the command chmod 755 program-name (or directory-name) to correct the problem.

Big Brother cannot find files or programs

Sometimes the operating system-specific program paths found in the installation might be wrong for a particular host's operating system. You can check and, as necessary, change the paths in the etc/bbsys.local file.  The variables contained in this file all have default values from the bbsys.OS file.  The original bbsys.local file is copied from the install/bbsys.OS file when you run the bbconfig program during installation, so keep track of your changes as they will automatically be overwritten if you run bbconfig again.