Home / user-259856

pkSML's questions

Martin Hope
pkSML
Asked: 2023-07-24 22:55:21 +0800 CST
  • 5

I've been searching for how to install two instances of nginx on my server without Docker/containerization, but I've found little results. How can this be done?

Possible use cases:

  • You have two applications and you don't want one to take down both if one crashes. (This is my scenario.)
  • You want to give an end user control of the nginx config and process/service. Since the processes are isolated from each another, they can be stopped/started & configured independently.
linux
Martin Hope
pkSML
Asked: 2022-04-03 16:01:12 +0800 CST
  • 0

I am in a conundrum. I need to extract data from a backup image I created previously. I thought I knew enough about LVM to get by... at least until now.

It seems my blunder is that I backed up an LVM snapshot volume rather than the whole volume group or disk partition. Most articles I've read imply that you've backed up the entire disk.

Here is my LVM layout:

xvda          202:0    0    10G  0 disk
└─xvda1       202:1    0    10G  0 part
  ├─lvm1-root 254:2    0   8.1G  0 lvm  /
  └─lvm1-swap 254:3    0   956M  0 lvm  [SWAP]

Here is my backup command:

lvcreate -L800M -s -n rootsnapshot /dev/lvm1/root
dd if=/dev/lvm1/rootsnapshot conv=sync,noerror bs=128K status=progress of=/path/to/snapshot.dd

I've mounted the DD file as /dev/loop0 on another machine.

Here is what I get:

# fdisk -l

Disk /dev/loop0: 8.1 GiB, 8657043456 bytes, 16908288 sectors
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes

# lvmdiskscan
  /dev/loop0 [       8.06 GiB]
  /dev/sda1  [      <7.00 GiB]
  /dev/sda5  [    1022.00 MiB]
  0 disks
  3 partitions
  0 LVM physical volume whole disks
  0 LVM physical volumes

# vgscan -v
  Reading all physical volumes.  This may take a while...
    No volume groups found.

Any idea how I might go about mounting this snapshot to retrieve data from it?

backup mount lvm loop-device
Martin Hope
pkSML
Asked: 2018-04-17 12:10:45 +0800 CST
  • 0

I have Dovecot 2.2.27 running successfully on Debian Stretch. I have some suspicious activity on my server and I want to log ALL raw IMAP communications sent/received by my server. Here is a .conf file in /etc/dovecot/conf.d/

service imap-login {
  executable = imap-login -R rawlogs
}

(Source: https://wiki.dovecot.org/Debugging/Rawlog )

Rawlog is recording when a user logs into my IMAP server (Pre-login rawlog). But that's when it stops. It doesn't record any communication after that. I want all IMAP communication received/sent for a connection (logging in, checking mailbox, downloading new mail, etc.)

Here's an example .in file:

1523905191.015557 1 LOGIN ***EMAIL ADDRESS HERE*** ***PASSWORD HERE***

And the .out file:

1523905190.806295 * OK [CAPABILITY IMAP4rev1 LITERAL+ SASL-IR LOGIN-REFERRALS ID ENABLE IDLE AUTH=PLAIN AUTH=LOGIN] Dovecot ready.

I added the following code to the afore-mentioned .conf file for post-login logging:

service imap {
  # tell imap to do post-login lookup using a socket called "imap-postlogin"
  executable = imap imap-postlogin
}

# The service name below doesn't actually matter.
service imap-postlogin {
  # all post-login scripts are executed via script-login binary
  executable = script-login /usr/lib/dovecot/rawlog

  # the script process runs as the user specified here (v2.0.14+):
#  user = $default_internal_user
  user = pksml
  # this UNIX socket listener must use the same name as given to imap executable
  unix_listener imap-postlogin {
  }
}

(Source: https://wiki.dovecot.org/PostLoginScripting )

It made no difference. Logging still only exists for authentication and nothing further, even when I'm checking the mail account. Am I doing something wrong, or is this not a capability of Dovecot? Thanks.

PS I have reloaded the Dovecot service after making .conf file changes.

PS #2: Mail users are virtual users, not real Linux system users

EDIT - Current .conf file:

import_environment = $import_environment DEBUG=1

service imap-login {
  executable = imap-login -R rawlogs
}

protocol imap {
  rawlog_dir = /var/run/dovecot/login/rawlogs
}

service imap {
  executable = imap imap-postlogin
}

service imap-postlogin {
  executable = script-login -d /usr/lib/dovecot/rawlog -I
  unix_listener imap-postlogin {
  }
}
linux
Martin Hope
pkSML
Asked: 2018-04-07 08:28:51 +0800 CST
  • 1

I've researched for an answer to this question and can't find anything.

I have Postfix running on Debian Stretch. I will soon be setting up Rspamd.

Here's the logic of what I want to have happen when spam arrives:

  • Postfix receives a message from the contact form on my website (via PHPMailer)
  • Postfix sends the message to Rspamd for processing
  • Rspamd flags the message as spam and adds appropriate headers
  • Postfix does NOT send the message to [email protected], but rather [email protected]

From what I gather, it seems that milters do not have this ability. How can this be accomplished?

email
Martin Hope
pkSML
Asked: 2015-09-29 15:42:27 +0800 CST
  • 1

I have a Broadcom based Linksys WRT160n router running DD-WRT build 14896. The LAN and wireless clients get a 192.168.1.xxx IP address. They are on br0. I have a guest network in the 192.168.2.xxx range running on br1. The guest network has internet access. I have no firewall/iptables rules declared.

Here's what I want to do: allow communication initiated from br0 to flow to br1. Disallow communication initiated from br1 to br0 (which seems to already be happening).

Right now, I can't ping or tracert from a 192.168.1.xxx host to 192.168.2.xxx host. I would like to be able to make that happen.

Is this something that can be done?

iptables