Summary:ASTERISK-01125: absence of cdr-csv directory gives no warning
Reporter:florian (florian)Labels:
Date Opened:2004-02-29 08:25:40.000-0600Date Closed:2004-09-25 02:54:44
Versions:Frequency of
Description:Seems like removing the cdr-csv directory and creating a new one is not allowed 'on the fly' - asterisk fails to log after this - and it does so SILENTLY. Considering the source code is very aware of cdr-files not being accessible (flushing everytime etc) this should considered baaaad...


A logcycle script moved cdr-csv and its contents elsewhere and created a new directory cdr-csv in its place. However, asterisk did not log to csv anymore, and it also gives no warning about this, as it would have done (I think) had the files become unaccessible.
Comments:By: James Golovich (jamesgolovich) 2004-03-02 14:45:29.000-0600

I'm not able to reproduce this.  I made a call to verify it was being written to the Master.csv file, then I moved the dir out of the way, mkdir a new cdr-csv, then made a new call and a new file was created containing the record.

Then I removed the cdr-csv and made a call to see what would happen and the following error was generated:
cdr_csv.c:219 csv_log: Unable to re-open master file /var/log/asterisk//cdr-csv//Master.csv

By: florian (florian) 2004-03-03 03:17:54.000-0600

Hmm, right - I just tried this again (cannot tinker too much on this box, production) and have to conclude you are right in this aspect:

At first, no Master.csv and related files were created on a box that has been running for some time. Stopping it entirely caused some havoc, so I rebooted and tried again. This time it worked properly. I have to conclude my bugreport is incorrect (or inaccurate at best) but I do remain confident there are some situations that can cause cdr-csv to not log without telling. I just have to try and figure out where and why. Will report back then. In the meantime - this bug can probably be closed.

By: James Golovich (jamesgolovich) 2004-03-03 03:36:31.000-0600

Closing this bug.  If you can come up with a scenario where it fails please reopen the bug