buffered-sql problem

[Alan DeKok]

On Jan 26, 2017, at 2:23 AM, Jouni Soini via Freeradius-Users <freeradius-users at lists.freeradius.org> wrote:
> 
> I was using radmin to log debug to a file and it seems the lines you mentioned is still written to radius.log
> if running freeradius -X gives more info I can try that too.

  It doesn't give *more* information, it gives timely information.

  If the logs are split between radius.log and another one... it's difficult to correlate the messages.  Possible, but difficult.

> these are the lines in radius.log:

  That shows normal processing of the detail.work file.

  Again, are you clear that writing to detail-20170125 does *not* mean that the packet is immediately read into the detail.work file?

  I just don't see how the server "misses" packets from the detail file.  The only evidence you've shown is the server writing to one file, and reading from another.  That's how it works.  If you wait a while, the server will finish the current detail.work, rename detail-20170125 to detail.work, and process that.

  If you see accounting packets being lost.. blame the NAS.  FreeRADIUS doesn't throw packets away.  The NAS does.

  Alan DeKok.