env_From incrementing

On 12/8/05 at 3:21 PM, vjs@calcite.rhyolite.com wrote:

>> What I'm seeing is that the env_From count is incrementing for
>> every unique, diferent message from the same sender.

>> # grep "  env_From:" $(grep -l kailacat msg.*)
>> msg.3HiJxG:   env_From: 070d67f5 175c2605 92be632c 27bff934      11
>> msg.3HiJxT:   env_From: 070d67f5 175c2605 92be632c 27bff934      12
>> msg.3HiJxf:   env_From: 070d67f5 175c2605 92be632c 27bff934       6
>> msg.3HiJxo:   env_From: 070d67f5 175c2605 92be632c 27bff934       7
>> msg.3HiJxv:   env_From: 070d67f5 175c2605 92be632c 27bff934       8
>> msg.3HiJxw:   env_From: 070d67f5 175c2605 92be632c 27bff934       9
>> msg.3HiJxz:   env_From: 070d67f5 175c2605 92be632c 27bff934      10
>> msg.3HiJyi:   env_From: 070d67f5 175c2605 92be632c 27bff934      13

> Where is the "incrementing"?  Have you checked the dates on those
> log files to find that they should be ordered by the last column?

# for i in  $(ls -t $(grep -l kailacat msg.*)) ; do
  grep " env_From:" $i ; done 
  env_From: 070d67f5 175c2605 92be632c 27bff934      13
  env_From: 070d67f5 175c2605 92be632c 27bff934      12
  env_From: 070d67f5 175c2605 92be632c 27bff934      11
  env_From: 070d67f5 175c2605 92be632c 27bff934      10
  env_From: 070d67f5 175c2605 92be632c 27bff934       8
  env_From: 070d67f5 175c2605 92be632c 27bff934       9
  env_From: 070d67f5 175c2605 92be632c 27bff934       7
  env_From: 070d67f5 175c2605 92be632c 27bff934       6

So it seems that each message has an env_From count one higher than
the last, except that the messages with counts of 8 and 9 are reversed
in the list above. Those two files have time stamps in the same minute
(and DCC time stamps in the same second), so ls -t lists them
alphabetically.

> Or are they messages each sent to 11, 12, 6, 7, 8, ... users?

None of these were addressed to more than one recipient.

> If each message is sent to one user, which DCC server are you using
> that records env_From checksums?

All of these have "X-DCC-NIET-Metrics" headers.

> If the server you are using is running with -Kenv_From, then it
> should count env_From values regardless of body checksum values.

OK.

> What `dccm -t` thresholds are you using?
> Are you using DCCM_REJECT_AT in /var/dcc/dcc_conf?

I'm not using a -t argument and am using DCCM_REJECT_AT in the config
file. It is set to 10, which I know is very low, but it's been that
way for a long time.

> Unless you are using `dccm -tALL,X` or `dccm -tenv_From,X`, whatever
> the DCC server says about the total for an env_From value should be
> ignored.

DCCM_ARGS in my dcc_conf is "-U userdirs -SHELO -x -r "<reject
message>" (that's not my real rejection message, I just skipped it for
clarity). So I'm not using a -t argument to dccm.

Thanks very much for your response.

-- John