On 05/09/12 13:05, Gerardo Herzig wrote:
Hi all. This is my intented scenario:

For all INCOMING mail:
- If count (for a given user@domain) reach, say, 10, put it on HOLD
- If count (for the same user@domain) reach, say, 20, DEFER it back

For that, i have configured 2 policies, like this one
sqlite>  select  * from policies;
ID|Name                  |Priority|Description           |Disabled
3 |Default Inbound       |10      |Default Inbound System|0
9 |Default Inbound Prio 5|5       |Higher priority       |0


And two quotas asociated with that policies:

sqlite>  select id,policyid,track,verdict,disabled from quotas;
ID|PolicyID|Track             |Verdict|Disabled
6 |3       |Sender:user@domain|HOLD   |0
11|9       |Sender:user@domain|DEFER  |0

The limits asociated with that quotas are:
sqlite>  select * from quotas_limits;
ID|QuotasID|Type        |CounterLimit|Disabled
7 |6       |MessageCount|3           |0
12|11      |MessageCount|5           |0

As im sending messages, both counters to increase, as expected:
sqlite>  select * from quotas_tracking;
QuotasLimitsID|TrackKey                |LastUpdate|Counter
12            |Sender:[email protected]|1336569907|1.99027777777778
7             |Sender:[email protected]|1336569907|1.99027777777778

When quota of id 6 reach its limits, it begins to HOLD (again, as
expected)

But...counter of quota of id 11 gets no more updated!! And that is *not*
what i want. I want to HOLD only *some* emails, if there is too much
mail from that user@domain, i want to DEFER (or REJECT maybe, im not
shure).


So....there is a way to get this thing done?

Your configuration looks good. You're right, lower prio's get matched first. What you say above should work as you expect.

Could you enable full debugging with all log detail and attach to a reply? Please don't exclude anything, I need to see the version of policyd you using along with each log line.

I honestly cannot see how you getting the behavior above from the source code. Even on a live system I cannot reproduce it using 2.0.11 or 2.1.x :) . I want to first verify against your logs that what you say above is indeed happening and that the ordering in the logs is either correct or incorrect... from there I can cross ref against the source and fix.

-N

Attachment: smime.p7s
Description: S/MIME Cryptographic Signature

_______________________________________________
Users mailing list
[email protected]
http://lists.policyd.org/mailman/listinfo/users

Reply via email to