On 1/29/2013 at 1:43 PM Brian Evans wrote: |On 1/29/2013 1:29 PM, Mike. wrote: |> |> On 1/29/2013 at 1:14 PM Brian Evans wrote: |> |> |On 1/29/2013 1:07 PM, Mike. wrote: |> |> I implemented the postscreen capability on a small MTA I run for |> |> friends and family. Once I got postscreen configuration producing |> the |> |> results I wanted, I soon tired of watching the detailed maillog to |> see |> |> how postscreen was operating. So I wrote a quick shell script to |> |> summarize the log file and give me an overview of how well |> postscreen |> |> is working. |> |> |> |> I offer the script to anyone who would like to use it. One company |> I |> |> worked for would not allow open source software into the company |> unless |> |> there was an explicit license on the software, so I put the BSD |> license |> |> on the script. |> |> |> |> You can download the script from here: |> |> http://archive.mgm51.com/sources/pslogscan.html |> |> |> |Fails without modification on my Gentoo mailserver: |> |Scanning /var/log/maillog |> |mktemp: too few X's in template mailqscan |> | |> |All "incoming" log records: 10121 |> |./pslogscan.sh: line 51: ${TmpFile}: ambiguous redirect |> | |> |Changing mailqscan to mailqscan.XXX works. |> | |> |Brian |> |> ============= |> |> |> Thanks for the feedback. |> |> I only run FreeBSD, so I figure there may be some minor issues like the |> one you mention when running on other OS's. |> |> |> |Also, your expressions don't count real postscreen numbers for connects |and rejects. |Take into account the following lines. | |Jan 28 12:47:57 mx1 postfix/error[19363]: 3Yvy410c1Mz8GKk: |to=<someu...@yahoo.com>, relay=none, delay=2332, delays=2331/1.2/0/0.07, |dsn=4.7.0, status=deferred (delivery temporarily suspended: host |mta7.am0.yahoodns.net[66.94.238.147] refused to talk to me: 421 4.7.0 |[TS01] Messages from xx.xx.xx.xx temporarily deferred due to user |complaints - 4.16.55.1; see http://postmaster.yahoo.com/421-ts01.html) |Jan 28 12:48:26 mx1 postfix/smtp[19336]: 3Yvy4D6lG7z8GL8: |to=<someu...@yahoo.com.com>, relay=none, delay=2350, |delays=2319/0.05/31/0, dsn=4.4.1, status=deferred (connect to |yahoo.com.com[216.239.120.187]:25: Connection timed out) | |Because of that, I have skewed numbers: |All "incoming" log records: 10187 |All "status=sent" log records: 7506 |All "status=deferred" log records: 3302 |rejected: -621 (-6%) | |It is not a simple math of "A minus B minus C" to find out how much |postscreen is rejecting in its current state. | |Brian
============= Yup. When there are a lot of deferrals, then things get complicated, requiring one to following individual messages through the process to eliminate multiple deferrals, etc., e.g., a single message "incoming" can get deferred many times leading to the numbers you cite. I wanted to keep things simple, so I made some compromises on the accuracy. I've been flipping back and forth between keeping the count of deferrals in there, or taking it out. Perhaps I should report the deferrals, but not count them in the expression.... hmmm...
