Statistics: Spam Scores shows negative values

G

ggzengel

Member
Jul 8, 2019
18
1
6
1618274150232.png

# pmgversion -v proxmox-mailgateway: 6.4-1 (API: 6.4-3/225ecb80, running kernel: 5.4.106-1-pve) pmg-api: 6.4-3 pmg-gui: 2.4-1 pve-kernel-5.4: 6.3-8 pve-kernel-helper: 6.3-8 pve-kernel-5.4.106-1-pve: 5.4.106-1 pve-kernel-5.4.103-1-pve: 5.4.103-1 pve-kernel-5.4.101-1-pve: 5.4.101-1 pve-kernel-5.4.78-2-pve: 5.4.78-2 pve-kernel-5.4.73-1-pve: 5.4.73-1 clamav-daemon: 0.102.4+dfsg-0+deb10u1 libarchive-perl: 3.3.3-1 libjs-extjs: 6.0.1-10 libjs-framework7: 4.4.7-1 libpve-apiclient-perl: 3.1-3 libpve-common-perl: 6.3-5 libpve-http-server-perl: 3.1-1 libxdgmime-perl: 0.01-5 lvm2: not correctly installed pmg-docs: 6.4-1 pmg-log-tracker: 2.1.9-1 postgresql-11: 11.11-0+deb10u1 proxmox-mini-journalreader: 1.1-1 proxmox-spamassassin: 3.4.5-1 proxmox-widget-toolkit: 2.4-9 pve-firmware: 3.2-2 pve-xtermjs: 4.7.0-3
 
Please post:
* a screenshot of the general Statistics page
* the output of: `pmgsh get /statistics/mail -starttime 1618178400 -endtime 1618264800`
* the output of: ` pmgsh get /statistics/spamscores -starttime 1618178400 -endtime 161826480`
(both outputs in CODE tags please)
 
1618306335279.png

Code: 
# pmgsh get /statistics/mail -starttime 1618178400 -endtime 1618264800
200 OK
{
   "avptime" : 9.98109970092773,
   "bounces_in" : 2,
   "bounces_out" : 0,
   "bytes_in" : 1328831,
   "bytes_out" : 2568994,
   "count" : 295,
   "count_in" : 294,
   "count_out" : 1,
   "glcount" : 1,
   "junk_in" : 279,
   "junk_out" : 0,
   "pregreet_rejects" : 248,
   "rbl_rejects" : 27,
   "spamcount_in" : 3,
   "spamcount_out" : 0,
   "spfcount" : 0,
   "viruscount_in" : 0,
   "viruscount_out" : 0
}

Code: 
# pmgsh get /statistics/spamscores -starttime 1618178400 -endtime 161826480
200 OK
[
   {
      "count" : 0,
      "level" : "0",
      "ratio" : 0
   },
   {
      "count" : 0,
      "level" : "1",
      "ratio" : 0
   },
   {
      "count" : 0,
      "level" : "2",
      "ratio" : 0
   },
   {
      "count" : 0,
      "level" : "3",
      "ratio" : 0
   },
   {
      "count" : 0,
      "level" : "4",
      "ratio" : 0
   },
   {
      "count" : 0,
      "level" : "5",
      "ratio" : 0
   },
   {
      "count" : 0,
      "level" : "6",
      "ratio" : 0
   },
   {
      "count" : 0,
      "level" : "7",
      "ratio" : 0
   },
   {
      "count" : 0,
      "level" : "8",
      "ratio" : 0
   },
   {
      "count" : 0,
      "level" : "9",
      "ratio" : 0
   },
   {
      "count" : 0,
      "level" : "10",
      "ratio" : 0
   }
]
 
Here the same day at UTC:

Code: 
# cat /etc/timezone
Etc/UTC

# pmgsh get /statistics/mail -starttime 1618185600 -endtime 1618272000
200 OK
{
   "avptime" : 1.45419465135408,
   "bounces_in" : 7,
   "bounces_out" : 9,
   "bytes_in" : 44981912,
   "bytes_out" : 71711384,
   "count" : 573,
   "count_in" : 464,
   "count_out" : 109,
   "glcount" : 0,
   "junk_in" : 280,
   "junk_out" : 0,
   "pregreet_rejects" : 248,
   "rbl_rejects" : 27,
   "spamcount_in" : 5,
   "spamcount_out" : 0,
   "spfcount" : 0,
   "viruscount_in" : 0,
   "viruscount_out" : 0
}

# pmgsh get /statistics/spamscores -starttime 1618185600 -endtime 1618272000
200 OK
[
   {
      "count" : 155,
      "level" : "0",
      "ratio" : 0.82010582010582
   },
   {
      "count" : 11,
      "level" : "1",
      "ratio" : 0.0582010582010582
   },
   {
      "count" : 18,
      "level" : "2",
      "ratio" : 0.0952380952380952
   },
   {
      "count" : 0,
      "level" : "3",
      "ratio" : 0
   },
   {
      "count" : 1,
      "level" : "4",
      "ratio" : 0.00529100529100529
   },
   {
      "count" : 0,
      "level" : "5",
      "ratio" : 0
   },
   {
      "count" : 1,
      "level" : "6",
      "ratio" : 0.00529100529100529
   },
   {
      "count" : 1,
      "level" : "7",
      "ratio" : 0.00529100529100529
   },
   {
      "count" : 1,
      "level" : "8",
      "ratio" : 0.00529100529100529
   },
   {
      "count" : 0,
      "level" : "9",
      "ratio" : 0
   },
   {
      "count" : 1,
      "level" : "10",
      "ratio" : 0.00529100529100529
   }
]
 
The email tells an other story:

Code: 
Proxmox Status Report - 2021-04-12 (lx-mailgateway)

System Status

Hostname    lx-mailgateway

Uptime    1 day 09:36

Version    pmg-api/6.4-3/225ecb80

Load    0.02

Memory    56.05%

Disk    17.03%

Incoming Mails (24 hours)

Incoming Mails    464

Junk Mails (60.34%)    280

Spam Mails (1.08%)    5

SPF rejects (0.00%)    0

Bounces (1.51%)    7

Virus Mails (0.00%)    0

Mail Traffic    42.898 MByte

Outgoing Mails (24 hours)

Outgoing Mails    109

Bounces (8.26%)    9

Mail Traffic    68.389 MByte

Spam Quarantine

Quarantine Size (MBytes)    12

Number of Mails    69

Average Size (Bytes)    181274

Average Spam Level    9
 
Stoiko Ivanov said:
* the output of: ` pmgsh get /statistics/spamscores -starttime 1618178400 -endtime 161826480`
Click to expand...
sorry - I mistyped - there is a 0 missing in the endtime:
`pmgsh get /statistics/spamscores -starttime 1618178400 -endtime 1618264800`
 
Code: 
# pmgsh get /statistics/spamscores -starttime 1618178400 -endtime 1618264800
200 OK
[
   {
      "count" : -14,
      "level" : "0",
      "ratio" : -0.7
   },
   {
      "count" : 11,
      "level" : "1",
      "ratio" : 0.55
   },
   {
      "count" : 18,
      "level" : "2",
      "ratio" : 0.9
   },
   {
      "count" : 0,
      "level" : "3",
      "ratio" : 0
   },
   {
      "count" : 1,
      "level" : "4",
      "ratio" : 0.05
   },
   {
      "count" : 0,
      "level" : "5",
      "ratio" : 0
   },
   {
      "count" : 1,
      "level" : "6",
      "ratio" : 0.05
   },
   {
      "count" : 1,
      "level" : "7",
      "ratio" : 0.05
   },
   {
      "count" : 1,
      "level" : "8",
      "ratio" : 0.05
   },
   {
      "count" : 0,
      "level" : "9",
      "ratio" : 0
   },
   {
      "count" : 1,
      "level" : "10",
      "ratio" : 0.05
   }
]
 
On a hunch - is the statistics still off for April 12. if you check it now?

The numbers do add up in the UTC-case..
 
Nice drift ...

> Do not trust any statistics you did not fake yourself.

Code: 
# date
Wed 14 Apr 2021 12:37:03 PM UTC

# pmgsh get /statistics/spamscores -starttime 1618178400 -endtime 1618264800
200 OK
[
   {
      "count" : -12,
      "level" : "0",
      "ratio" : -0.545454545454545
   },
   {
      "count" : 11,
      "level" : "1",
      "ratio" : 0.5
   },
   {
      "count" : 18,
      "level" : "2",
      "ratio" : 0.818181818181818
   },
   {
      "count" : 0,
      "level" : "3",
      "ratio" : 0
   },
   {
      "count" : 1,
      "level" : "4",
      "ratio" : 0.0454545454545455
   },
   {
      "count" : 0,
      "level" : "5",
      "ratio" : 0
   },
   {
      "count" : 1,
      "level" : "6",
      "ratio" : 0.0454545454545455
   },
   {
      "count" : 1,
      "level" : "7",
      "ratio" : 0.0454545454545455
   },
   {
      "count" : 1,
      "level" : "8",
      "ratio" : 0.0454545454545455
   },
   {
      "count" : 0,
      "level" : "9",
      "ratio" : 0
   },
   {
      "count" : 1,
      "level" : "10",
      "ratio" : 0.0454545454545455
   }
]
 
There is one more error:

On 11-04 there was 1 outgoing mail, but in global graphs it shows more.
On 12-04 there was more outgoing mail, but in global graphs it shows only one outgoing mail.

I think the value tables under global statistics is one day behind.

1618418548296.png
 
To avoid creating a new thread, we're seeing the same negative Score 0 reporting figure, which has a knock-on to the other score percentages.

Screenshot inline, which is from PMG 6.4-3. Did you ever get any further with potential "why"s? We've seen some emails scored negatively.

PMG-NegativeScore0-20210606.png

Hoping to hear back with many thanks!
 
linux said:
To avoid creating a new thread, we're seeing the same negative Score 0 reporting figure, which has a knock-on to the other score percentages.
Click to expand...

Following on from this, since the above screenshot we've deployed inbound filtering to PMG as well, which has reduced/improved the problem.

PMG-NegScore-Improved20210607.png

It's earlier in the day relative to when we took the screenshot yesterday, though the percentages are far more sane.

Very curious if anyone/Proxmox has any clues here as to what's causing the behaviour behind-the-scenes?

Thank you!
 
You must log in or register to reply here.
Top Bottom