My forum search was not fruitful. I have had LinHES 6.2 working for a few months. Yesterday, all recorded TV shows disappeared from the list. I rebooted the machine and that’s when the trouble really started. I am stuck in the Database Configuration screens.

The reboot sequence is different now:

1) it asks for your preferred language.
2) Message: No UPnP backends found
3) Database Configuration pp. 1 of 2
“Myth could not connect to the database. Please verify your database settings”:

• Hostname = localhost
• Ping test Server = checked
• Port = blank (blank is default = 3306)
• Database name = mythconverg
• User: mythtv
• Password: mythtv

Database Configuration pp. 2 of 2

Settings:
• Use Custom Identifier for frontend preferences = Unchecked
• Enable Database Server wakeup = Unchecked

I have attempted to login to mySQL from the command line:


QUESTIONS
==========
1) If you have seen this problem before (database logon?) , what is the trouble shooting procedure? I think this is DB logon problem but not 100% sure.
2) What happened? I would like to prevent this from happening again.

Thanks in advance for actionable insight.

_________________
R7.3: 0.22.20091023-1, Hauppauge PVR-500 (Philips FQ1236A MK4), Gigabyte Gigabyte EG45M-UD2H, E5200 2.4Ghz, 2GB RAM, NVIDIA GEFORCE 256MB

Make sure that "localhost" is one word (no spaces), or use the loopback address 127.0.0.1 that it should be mapped too.

Other things to check:

- Use pgrep to see if the db server is running.


- Check that localhost is mapped to the right address


- Check your network config


- Well, that's enough to get you started.

I ran into a problem after installing R6 where the config was trying to use my host name (black2), but that was bound to the ethernet (eth0) rather than the loopback (lo) device.

Thank you for the very specific and actionable guidance.

As you can see my hostname is mythtv (not very original). I verfied that that localhost was one word with (no space). Tried replacing the string with 127.0.0.1. No joy. Replaced 127.0.0.1 with original hostname (mythtv).

System indicates that credentials \ PW are problematic. I am unable to logon into mySQL from the command line. I am hoping to avoid a fresh install.

As prescribed: here are the test results.

_________________
R7.3: 0.22.20091023-1, Hauppauge PVR-500 (Philips FQ1236A MK4), Gigabyte Gigabyte EG45M-UD2H, E5200 2.4Ghz, 2GB RAM, NVIDIA GEFORCE 256MB

try the lan IP for your mythtv box also check to see how much room you have on your root partition my whole system went haywire when I ran out of room.

_________________
My System

@mattbatt: Good call. I do not know what is a reasonable amount of space would be, however, this is my test. If there is another test, please let me know the syntax and will post ther results.



97% of the partition one is used. This is the first time I've checked this so the question is what is a reasoable range. If this diskspace is the issue, how did you resolve this problem?

_________________
R7.3: 0.22.20091023-1, Hauppauge PVR-500 (Philips FQ1236A MK4), Gigabyte Gigabyte EG45M-UD2H, E5200 2.4Ghz, 2GB RAM, NVIDIA GEFORCE 256MB

This looks like your problem. The start up script is running, but the DB server process is not.. There should be a mysqld process. You need to figure out why that server isn't starting properly.

Try running the server using the command shown in my pgrep output above, but adding the "--verbose" option. That may provide some clues as to what the problem is.

OK, I am not complaining because it has somehow booted to a normal screen (there was a brief power failure this morning). I can now see the programs that were recorded.

I ran some tests again



@tjc: please advise the syntax of the verbose command you would like executed. Thanks.



I am not sure what was wrong with the database. Any command syntax to probe the database to test theories are appreciated.

_________________
R7.3: 0.22.20091023-1, Hauppauge PVR-500 (Philips FQ1236A MK4), Gigabyte Gigabyte EG45M-UD2H, E5200 2.4Ghz, 2GB RAM, NVIDIA GEFORCE 256MB

Hmmm... DB server eems to be running now, but your space in the root partition is _really_ low. At a guess you ran out of space while the DB was trying to do some update and "bad stuff" happened.

Can you login as mythtv? Try this (default pw is the same as the user name):


If that works try running one of the DB repair and optimization scripts to see what they complain about. Either optimize_db.sh or optimize_mythdb.pl should work.

We've see other people lose the mysql management schema which contains all the passwords and other administrative data for the DB. There are ways to fix that up but they're moderately involved.

I was not successful logging in with my first try. However, I have had success loggin in the past using the mythtv username to create an ERD.



I did a little homework (Google is my friend) and found that I could login if by appending -p to the command:



OK so now let's extend this to mythconverg database.



Before I run either optimize_db.sh or optimize_mythdb.pl experiment, should we expect that the diskspace be reduced? or is the purely a diagnostic? If you have any recommended syntax for running either command that would be appreciated, as it would be best to get it right on the first try. I have no experience running a .pl script and syntax help be appreciated. I think that the syntax for foo.sh is:



Confirmation before I execute would be appreciated.

As you can see I am up from 97% (yesterday) to 98% (today). Yikes Thank you.

_________________
R7.3: 0.22.20091023-1, Hauppauge PVR-500 (Philips FQ1236A MK4), Gigabyte Gigabyte EG45M-UD2H, E5200 2.4Ghz, 2GB RAM, NVIDIA GEFORCE 256MB

If the script is set up properly (and those are) you should be able to just run them by typing the name.

It looks like the options out of the box with R6 are optimize_mythdb.py and optimize_mythdb.pl. Both appear to do pretty much the same thing.