Re: [hylafax-users] ratio DONE/FAILED

[Lee Howard <faxguy@xxxxxxxxxxxxxxxx>]

Unless I specify otherwise, the error rates that I publish are based on 
numbers coming from recvstats and xferfaxstats.  I take the total number 
of errors shown and divide it by the total number of *pages* sent.  
(Using pages instead of calls or jobs makes a more meaningful result, in 
my opinion.)  On well-tuned installations receive rates are well below 
1%, often 0.1%.  On the send side it varies quite a bit.  If you have a 
user who constantly is dialing the wrong number then the error rates go 
up.  If you repeatedly send to a problematic receiver then the rates go 
up.  So sometimes I have been accustomed to pulling out "no carrier", 
"no answer", and "busy" errors from the statistical calculations.  But 
in no way should error rates be at 30% or even 10%.  Generally speaking 
your users are going to keep you aware of what an "acceptable" or 
"normal" sending error rate is - if you send them back failed job 
notification messages.  I use NOTIFY_FAXMASTER=errors always so that I 
often can be proactive about fixing issues before a user complains.  In 
sending situations I expect to normally see a 3% to 5% error rate... 
unless I have some extenuating circumstances (like a user always 
misdialing).  If you always send to the same receivers over and over and 
the pool is not random, then you can expect your error rates in sending 
to be closer to 1%.

Lee.


Moritz Winterberg wrote:

> you should take in consideration that my error rate does include all
> possible errors.
> Also such as Busy destination, wrong number, no answer and destination
> being voice number instead of fax.
> Since I'm running an end user system with about 5000 customers using
> it. Those errors obviously occur quite often.
> 500 calls means almost 500 different destinations here.
>
>
> On 5/4/07, Andrew Rinaldi <andrew.rinaldi@xxxxxxxxxxxx> wrote:
>
> > At some point in the past Lee published some error rates that he'd 
> > seen on
> > HylaFAX systems that he administers.  If I remember correctly, in the 
> > 'real
> > world' he reported error rates of approximately 0.5% if you included
> > operator errors such as "hitting the stop button". 

____________________ HylaFAX(tm) Users Mailing List _______________________
 To subscribe/unsubscribe, click http://lists.hylafax.org/cgi-bin/lsg2.cgi
On UNIX: mail -s unsubscribe hylafax-users-request@xxxxxxxxxxx < /dev/null
 *To learn about commercial HylaFAX(tm) support, mail sales@xxxxxxxxx*