login: root

DirectAdmin Modsecurity changes in Apache

Wed, 31 Dec 2025 13:39:05 +0200

DirectAdmin 1.691 release has modsecurity part refactored. It uses a different method to enable the modsecurity for the apache webserver.

the httpd-modsecurity.conf is now included with the httpd-modsecurity-enable.conf that gets generated dynamically, not with the httpd-phpmodules.conf as before
httpd-modsecurity-enable.conf is always included by the httpd.conf. That file is empty when modsec is disabled, and filled with includes when enabled.
libxml2.so is not used in apache configs anymore as it’s not needed for modsec in 2.4.

If you receive an error like this:

Exim Logging Port and Plaintext Connections

Fri, 23 May 2025 15:09:46 +0300

DirectAdmin recently has enforced stricter rules for sending mail (changelog).

The long story short is that starting with the DirectAdmin version 1.676:

Port 25 cannot be used for the mail submission. It’s dedicated for server to server comunication.
Authentication via port 587 works only when using StartTLS.

The changelog also provides a way to revert the full or partial functionality to the previous one in case the impact is bigger than expected.

However, the change is strongly adviced. But before jumping to the stricter mode, you may want to collect some stats to visualize how much users are actually using the forbidden settings.

Directadmin With a Mail Proxy

Thu, 22 May 2025 14:10:01 +0300

Mail proxy setup to have a common mail.domain.tld that proxies IMAP and SMTP connections to the appropriate DA servers.

mail.server.tld
 ├─da1.server.tld
 ├─da2.server.tld
 └─da3.server.tld

Directadmin Old Cyphers for Mail

Wed, 21 May 2025 10:32:34 +0300

If a user insists on maintaining support for legacy systems to enable connections using older SSL methods, one option is to set ssl_configuration=old in the options.conf file of custombuild. However, this setting changes the SSL ciphers for all web-related services as well. Alternatively, it is possible to downgrade the ciphers exclusively for mail services without affecting the web services.

DirectAdmin x509: certificate has expired

Mon, 24 Oct 2022 19:45:21 +0000

1
2

License check failure See the Debug Guide
Reason: request failed: Post "https://licensing.directadmin.com/start": x509: certificate has expired or is not yet valid: current time 2022-10-26T11:51:22+03:00 is after 2021-09-30T14:01:15Z

The issue is likely due to server CA root certificates being too old.

If you are using Centos 7 which is still supported – you should still be able to update your OS.

Check what is your redhat version with:

Fixing error “firmware: failed to load libertas/sd8688_helper.bin (-2)”

Tue, 14 Feb 2017 11:46:48 +0000

After debian install onto guruplug, I’ve noticed few errors:

[ 10.445696] libertas_sdio mmc0:0001:1: firmware: failed to load libertas/sd8688_helper.bin (-2)
[ 10.454505] libertas_sdio mmc0:0001:1: Direct firmware load for libertas/sd8688_helper.bin failed with error -2
[ 10.605080] libertas_sdio mmc0:0001:1: firmware: failed to load sd8688_helper.bin (-2)
[ 10.613113] libertas_sdio mmc0:0001:1: Direct firmware load for sd8688_helper.bin failed with error -2
[ 10.622488] libertas_sdio: failed to find firmware (-2)
[ 10.769048] btmrvl_sdio mmc0:0001:2: firmware: failed to load mrvl/sd8688_helper.bin (-2)
[ 10.777317] btmrvl_sdio mmc0:0001:2: Direct firmware load for mrvl/sd8688_helper.bin failed with error -2
[ 10.786956] Bluetooth: Failed to download helper!
[ 10.791706] Bluetooth: Downloading firmware failed!

Also, neither Bluetooth nor Wifi were working.

Installing Debian on GuruPlug Server Plus

Mon, 13 Feb 2017 19:43:23 +0000

Let’s hookup our GuruPlug JTAG board as described in GuruPlug serial connection post.

Latest debian testing images for GuruPlug may be found here.

Boot up the device:

=> usb start
starting USB...
USB0:   USB EHCI 1.00
scanning bus 0 for devices... 4 USB Device(s) found
       scanning usb for storage devices... Device NOT ready
   Request Sense returned 02 3A 00
3 Storage Device(s) found
=> usb part
## Unknown partition table type 0


Partition Map for USB device 1  --   Partition Type: DOS

Part    Start Sector    Num Sectors     UUID            Type


Partition Map for USB device 2  --   Partition Type: DOS

Part    Start Sector    Num Sectors     UUID            Type
  1     2048            7852032         77e6cfe3-01     0b

## Unknown partition table type 0

## Unknown partition table type 0

## Unknown partition table type 0

## Unknown partition table type

Load images to a NAND temporary location

Updating u-boot on GuruPlug Server PLUS

Fri, 03 Feb 2017 20:57:08 +0000

After long post starvation period GuruPlug got chance to be used again.
It seems logical, that all the software needs updated. I’ve checked my old post, and it seems like it has some parts missing, and those links with u-boot images doesn’t have any recent update.

So I’m here to fix that part.
As it’s a follow up to the old uboot install post I won’t get into details.

Don’t forget the disclaimer:
https://loginroot.com/disclaimer/
Please, keep in mind, that rewriting U-Boot erases all the environment variables including the ones that contain Your boot parameters, so Your existing OS will become unbootable!
The new OS install will be covered in the next post.

Getting rid of “kernel:nf_ct_ftp” messages

Wed, 27 Jan 2016 12:47:56 +0000

Sometimes firewall drops nf_ct_ftp messages to console and disturbs the work during increased amount of traffic to FTP port.

Message from syslogd@web1 at Jan 27 12:28:40 ...
 kernel:nf_ct_ftp: dropping packetIN= OUT=eth0 SRC=94.23.15.229 DST=36.252.2.15 LEN=53 TOS=0x10 PREC=0x00 TTL=64 ID=62620 DF PROTO=TCP SPT=21 DPT=61868 SEQ=256869292 ACK=3387663834 WINDOW=14600 RES=0x00 ACK PSH FIN URGP=0

Message from syslogd@web1 at Jan 27 12:28:40 ...
 kernel:nf_ct_ftp: dropping packetIN= OUT=eth0 SRC=94.23.15.229 DST=36.252.2.15 LEN=53 TOS=0x10 PREC=0x00 TTL=64 ID=62621 DF PROTO=TCP SPT=21 DPT=61868 SEQ=256869292 ACK=3387663834 WINDOW=14600 RES=0x00 ACK PSH FIN URGP=0

Message from syslogd@web1 at Jan 27 12:28:41 ...
 kernel:nf_ct_ftp: dropping packetIN= OUT=eth0 SRC=94.23.15.229 DST=36.252.2.15 LEN=53 TOS=0x10 PREC=0x00 TTL=64 ID=13439 DF PROTO=TCP SPT=21 DPT=32394 SEQ=1981989941 ACK=3962951258 WINDOW=14600 RES=0x00 ACK PSH FIN URGP=

Those messages may be redirected by creating a file /etc/rsyslog.d/iptables.conf with contents:

The service ‘freshclam’ on server server.loginroot.com is currently down (Debian 8)

Wed, 06 Jan 2016 07:45:17 +0000

If You are getting mails from DA about:

`1`	`The service 'freshclam' on server server.loginroot.com is currently dow`

And status after freshclam restart:

# service freshclam status
● freshclam.service - ClamAV updater daemon
   Loaded: loaded (/etc/systemd/system/freshclam.service; enabled)
   Active: failed (Result: timeout) since Wed 2016-01-06 08:29:31 CET; 2s ago
  Process: 13369 ExecStart=/usr/local/bin/freshclam -d (code=exited, status=0/SUCCESS)

Jan 06 08:28:01 server.loginroot.com freshclam[13370]: freshclam daemon 0.99 (OS: linux-gnu, ARCH: x86_64, CPU: x86_64)
Jan 06 08:28:01 server.loginroot.com freshclam[13370]: ClamAV update process started at Wed Jan  6 08:28:01 2016
Jan 06 08:28:01 server.loginroot.com freshclam[13370]: main.cvd is up to date (version: 55, sigs: 2424225, f-level: 60, builder: neo)
Jan 06 08:28:01 server.loginroot.com freshclam[13370]: daily.cld is up to date (version: 21229, sigs: 1786793, f-level: 63, builder: neo)
Jan 06 08:28:01 server.loginroot.com freshclam[13370]: bytecode.cvd is up to date (version: 270, sigs: 46, f-level: 63, builder: shurley)
Jan 06 08:28:01 server.loginroot.com freshclam[13370]: --------------------------------------
Jan 06 08:29:31 server.loginroot.com systemd[1]: freshclam.service start operation timed out. Terminating.
Jan 06 08:29:31 server.loginroot.com freshclam[13370]: Update process terminated
Jan 06 08:29:31 server.loginroot.com systemd[1]: Failed to start ClamAV updater daemon.
Jan 06 08:29:31 server.loginroot.com systemd[1]: Unit freshclam.service entered failed state.

It’s because service waits for a pid to appear in /var/run/clamd, but directory doesn’t exist, and it has insufficient permissions to create it.

Yubikey

Fri, 25 Dec 2015 20:30:53 +0000

1. What is Yubikey?

A YubiKey is a small hardware device that offers two-factor authentication with a simple touch of a button. YubiKeys are built strong enough for the largest enterprises, while remaining simple enough for anyone to use. The YubiKey NEO offers both contact (USB) and contactless (NFC, MIFARE) communications. YubiKeys support FIDO U2F, Yubico-OTP, OATH-OTP, OATH-HOTP, OATH-TOTP, OpenPGP, and PIV, and one security key can support an unlimited number of applications without the need for drivers, client software, or batteries.

WordPress visual html editor missing (only Text mode)

Fri, 25 Dec 2015 19:06:25 +0000

It was a pitty, when richer text format was required, but none of the features were visible.
And when installing MCEtiny, CKeditor and other plugins didn’t do anything for that, it brought me an idea that some plugin may be interfering with the latest wordpress version.

Fortunately answer was incredibly simple. It seems like WP users have a setting of ‘visual’ editor, that was disabled.

To reenable Visual editor, go to user settings, and find the place:

“bash: scp: command not found” despite it exists

Wed, 09 Dec 2015 10:03:44 +0000

If You receive the error something like, despite the scp exist:

nsc-book:~ nsc$ scp [email protected]:/root/filename ~/filename
bash: scp: command not found
nsc-book:~ nsc$ which scp
/usr/bin/scp

It may look silly, but please check if the destination server has that package too ;)
I haven’t expected a CentOS server not having this simple package by default.

git error “server certificate verification failed”

Tue, 08 Dec 2015 18:45:30 +0000

I’ve got this message on Ubuntu 12.04. Other servers seemed to work fine.

root@server:/tmp# git clone https://git.kolab.org/diffusion/RPK/roundcubemail-plugins-kolab.git
Cloning into 'roundcubemail-plugins-kolab'...
error: server certificate verification failed. CAfile: /etc/ssl/certs/ca-certificates.crt CRLfile: none while accessing https://git.kolab.org/diffusion/RPK/roundcubemail-plugins-kolab.git/info/refs
fatal: HTTP request failed

The fix:
Append CA root of that cert to file: /etc/ssl/certs/ca-certificates.crt

OR
just to disable sslverify for git globally (it may be reenabled afterwards):

Cracking the RSA keys (Part 2 – generating the private key)

Mon, 23 Nov 2015 15:23:10 +0000

Previous part:
Part 1 - Cracking the RSA keys (Part 1 - getting the private exponent)

PART 2
For a private key generation, we need to create an asn1parse.txt file with the contents of:

asn1=SEQUENCE:rsa_key

[rsa_key]
version=INTEGER:0
modulus=INTEGER:0
pubExp=INTEGER:0
privExp=INTEGER:0
p=INTEGER:0
q=INTEGER:0
e1=INTEGER:0
e2=INTEGER:0
coeff=INTEGER:0

Explanation:
Modulus - it’s the product of two prime keys. It’s visible in public key.
pubExp - public exponent, it’s also visible in public key.
privExp - private exponent (d) that was calculated in previous post.
p - first prime number
q - second prime number
So we have all these from a previous post.

Cracking the RSA keys (Part 1 – getting the private exponent)

Mon, 23 Nov 2015 15:17:30 +0000

The whole idea of the RSA private key is the hardness of factorisation of two very large prime numbers.
That’s why recommended RSA keys are >2048bit long.
I won’t get into RSA details itself. If You need any info, it’s here: WIKI For a sake of demonstration, 256bit public_key will be used. With the current hardware that we have these days, it’s very easy crackable.

-----BEGIN PUBLIC KEY-----
MDwwDQYJKoZIhvcNAQEBBQADKwAwKAIhALDxlk/H4PXkJ2ERM3PZmXB5cH0ApFr+
IrDuluL/kQIzAgMBAAE=
-----END PUBLIC KEY----

Lets convert it to a more ‘mathematical’ expression:

Add a separate SSL certificate for a subdomain in Directadmin

Mon, 27 Jul 2015 18:40:38 +0000

Directadmin doesn’t have possibility to add SSL certificate for a subdomain, unless You add it as a separate domain in domain control.

You may also use separate SSL certificates by creating a custom httpd (or nginx, but it would require to change the CAROOT line) file in user configuration dir i.e. loginroot.com.cust_httpd (or loginroot.com.cust_nginx if it’s nginx).

 >|*if SUB="subdomain"|
	|?KEY=/usr/local/directadmin/data/users/loginroot/domains/subdomain.loginroot.com.key|
	|?CERT=/usr/local/directadmin/data/users/loginroot/domains/subdomain.loginroot.com.cert|
	|?CAROOT=SSLCACertificateFile /usr/local/directadmin/data/users/loginroot/domains/subdomain.loginroot.com.cacert|
|*endif

Wrong LSAPI header type: 6 instead of expected 7

Thu, 09 Jul 2015 17:57:17 +0000

If You are getting this error after update:

[Thu Jul 09 13:32:08.465065 2015] [lsapi:error] [pid 467230:tid 140242392512256] [client 10.3.13.37:65135] [host loginroot.com] Backend error on sending request(GET / HTTP/1.1); uri(/) content-length(0) (lsphp is killed?): Wrong LSAPI header type: 6 instead of expected 7

That’s because cloudlinux developers modified their lsapi code, and didn’t add variable default value. Very nice guys…

The fix:

Just add the line to /etc/httpd/conf/extra/mod_lsapi.conf (Or other lsapi configuration file in case Your server is not a DA server):

DBD::mysql perl module hasn’t been fully installed

Wed, 24 Jun 2015 11:04:20 +0000

DADA mail error:

[Wed Jun 24 11:35:49 2015] mail.cgi: install_driver(mysql) failed: Can't locate DBD/mysql.pm in @INC (@INC contains: ../../ ../../perllib /sw/lib/perl5/5.8.6/darwin-thread-multi-2level /sw/lib/perl5 ../../DADA/perllib ../../../ ../../DADA ../../../DADA/perllib ../../../perllib ./ ../ ./../../DADA /home/login/domains/loginroot.com/public_html/mailinglist/DADA/perllib /home/login/domains/loginroot.com/public_html/mailinglist /usr/local/lib64/perl5 /usr/local/share/perl5 /usr/lib64/perl5/vendor_perl /usr/share/perl5/vendor_perl /usr/lib64/perl5 /usr/share/perl5 . /home/login/perl5/lib/perl5 /home/login/perl5/lib/perl5/x86_64-linux-thread-multi /home/login/perllib /home/login/perl/home/login/domains/loginroot.com/public_html/mailinglist/DADA/perllib /home/login/perl/home/login/domains/loginroot.com/public_html/mailinglist /home/login/perl/usr/local/lib64/perl5 /home/login/perl/usr/local/share/perl5 /home/login/perl/usr/lib64/perl5/vendor_perl /home/login/perl/usr/share/perl5/vendor_perl /home/login/perl/usr/lib64/perl5 /home/login/perl/usr/share/perl5 /home/login/perl.) at (eval 20) line 3, <CONFIG> line 1.
[Wed Jun 24 11:35:49 2015] mail.cgi: Perhaps the DBD::mysql perl module hasn't been fully installed,
[Wed Jun 24 11:35:49 2015] mail.cgi: or perhaps the capitalisation of 'mysql' isn't right.
[Wed Jun 24 11:35:49 2015] mail.cgi: Available drivers: DBM, ExampleP, File, Gofer, Proxy, Sponge.
[Wed Jun 24 11:35:49 2015] mail.cgi: at /DADA/App/DBIHandle.pm line 16

The fix

wsgi_python.h missing

Sat, 20 Jun 2015 19:42:32 +0000

/var/www/build/libtool --silent --mode=compile gcc -std=gnu99 -prefer-pic   -DLINUX -D_REENTRANT -D_GNU_SOURCE -g -O2 -pthread -I/usr/include/apache  -I/usr/include/apache   -I/usr/include/apache  -g -O2 -I/usr/include/python2.7 -DNDEBUG  -c -o src/server/mod_wsgi.lo src/server/mod_wsgi.c && touch src/server/mod_wsgi.slo
In file included from src/server/mod_wsgi.c:22:0:
src/server/wsgi_python.h:24:20: fatal error: Python.h: No such file or directory
compilation terminated.
apxs:Error: Command failed with rc=65536
.
make: *** [src/server/mod_wsgi.la] Error 1

Solution:
Debian/Ubuntu:

`1`	`apt-get install python-de`

Centos/RHEL:

Yubikey 2 factor authentication for wordpress (U2F)

Sat, 06 Jun 2015 20:32:15 +0000

At last my [Yubikey][1] arrived!
Yubikey-edge to be more precise.

First thing I wanted to do is to enable 2factor authentication for this blog :)

I’ve noticed that yubico has an old info about wordpress plugin for yubikey Plugin seems to work fine, but it uses Yubico Web service API for authentication, despite that U2F protocol’s technical specifications were published by google pretty long time ago.
After some time has passed, wordpress 2nd auth plugin with U2F support was released.

The host at this IP address is infected with the CryptPHP PHP malware.

Sun, 23 Nov 2014 13:49:49 +0000

Have anyone else got the same problem with cbl.abuseat.org??

Full message is here:

The host at this IP address is infected with the CryptPHP PHP malware.

CryptoPHP is a threat that uses backdoored Joomla, WordPress andn Drupal themes and plug-ins to compromise webservers on a large scale. More information about this threat can be found on the referenced link below.

Fox-IT: CryptoPHP - Analysis of a hidden threat inside popular content management systems
Attackers Using Compromised Web Plug-Ins in CryptoPHP Blackhat SEO Campaign
This infection almost certainly means that the infected web site has used pirated plugins from the nulledstylez.com, dailynulled.com sites or some other site that specializes in providing "nulled" (pirated) software. Fox-IT's research has shown that every pirated theme or plug-in on these two sites has been infested with the cryptophp malware.

This was detected by a TCP connection from 1.1.1.1 on port 44874 going to IP address 192.42.116.41 (the sinkhole) on port 80.

The botnet command and control domain for this connection was "carandfly.net".

Behind a NAT, you should be able to find the infected machine by looking for attempted connections to IP address 192.42.116.41 or host name carandfly.net on any port with a network sniffer such as wireshark or by configuring the router to block and log such connections. Equivalently, you can examine your DNS server or proxy server logs to references to 192.42.116.41 or carandfly.net. See Advanced Techniques for more detail on how to use wireshark - ignore the references to port 25/SMTP traffic - the identifying activity is NOT on port 25.

This detection corresponds to a connection at 2014-11-23 10:09:06 (GMT - this timestamp is believed accurate to within one second).

Note that in the description below, findbot can sometimes yield a lot of false positives. In this case, you are probably looking for references to "social.php", so grepping the output of findbot for "social.php", may work better.

Additionaly, they provide a tool findbot.pl that seems to find more false positives than real threats, so basically it’s a huge time waster.
Output grepping for social.php wasn’t a big help either (If You tried that, You would know what I mean :) )
As we want ‘blaclisting’ cause to be found ASAP, that script does it’s job pretty crappy. Client is not going to wait for a week till You cross out thousands of false positives.

Cacti Allowed memory size of 123456 bytes exhausted

Mon, 17 Nov 2014 20:59:24 +0000

If You get such error on after installing cacti, and no mater how big memory_limit:

[Mon Nov 17 22:41:49.267134 2014] [:error] [pid 24152] [client 78.58.224.49:41062] PHP Fatal error:  Allowed memory size of 2147483648 bytes exhausted (tried to allocate 130968 bytes) in /usr/share/cacti/lib/adodb/adodb.inc.php on line 833

Please make sure if You imported the cacti sql file.
My .sql file location was:
_/usr/share/doc/cacti-0.8.8b/cacti.sql
_

You may easily import the database like:

`1`	`cat /usr/share/doc/cacti-0.8.8b/cacti.sql \| mysql -u username database -p`

How-to: open/close the luks partition

Wed, 05 Nov 2014 21:26:13 +0000

The title is pretty self-explanatory :)

To open the luks partition:

`1`	`cryptsetup -v luksOpen /dev/local/home home_crypte`

And to close it:

`1`	`cryptsetup -v luksClose home_crypte`

qemu-kvm: -chardev pty,id=charserial0: Failed to create chardev

Tue, 28 Oct 2014 22:55:17 +0000

This error was received on CentOS 7, at the beginning of the install command:

`1`	`ERROR internal error: process exited while connecting to monitor: qemu-kvm: -chardev pty,id=charserial0: Failed to create charde`

The fix is to edit /etc/fstab find the devpts line and change it to:

`1`	`devpts /dev/pts devpts gid=5,mode=620 0 0`

Run remount command

`1`	`mount -o remount /dev/pts`

And it should be enough

Booting KVM VM from CD image

Tue, 28 Oct 2014 19:06:38 +0000

You may list Your all KVM VM with command:

[root@kvm3 qemu]# virsh list --all
 Id    Name                           State
----------------------------------------------------
 -     vm1.server.tld               shut off

Edit the config file:

`1`	`[root@kvm3 qemu]# virsh edit vm1.server.tld`

Add the following lines to the ‘devices’ part:

<disk type='file' device='cdrom'>
  <driver name='qemu' type='raw'/>
  <source file='/tmp/CentOS-7.0-1406-x86_64-Minimal.iso'/>
  <target dev='hdb' bus='ide'/>
  <readonly/>
  

<address type='drive' controller='0' bus='1' unit='0' />
</disk>

Script: populating the domainips file for the exim automatically (DirectAdmin)

Thu, 23 Oct 2014 13:59:41 +0000

Script for populating /etc/virtual/domainips records for the exim.
Exim configuration change is described [here][2]
It takes IP that is assigned to that domain in Directadmin and adds it to the domainsip file.

Here is the script (Please note: it always overwrites the domainips file!!!):

#!/bin/sh

LIST=`/bin/cat /etc/virtual/domainowners | tr -d ' '`
echo -n '' > /etc/virtual/domainips

for LINE in $LIST; do
	USER=`echo $LINE | cut -d: -f2`
	DOMAIN=`echo $LINE | cut -d: -f1`
	if [ -f "/usr/local/directadmin/data/users/$USER/domains/$DOMAIN.conf" ]; then
		IP=`/bin/grep 'ip=' /usr/local/directadmin/data/users/$USER/domains/$DOMAIN.conf | cut -f2 -d=`
		[ -z $IP ] || echo "$DOMAIN: $IP" >> /etc/virtual/domainips
	fi
done

exit 0;

note: You may add this script to cron, for automatic ‘domainips’ updates.

How-to: Sending emails using IP address per domain in exim (Directadmin)

Thu, 23 Oct 2014 13:56:55 +0000

To be able to send mails via different IP per domain, You need to do the following changes in the /etc/exim.conf file:

Find the ‘remote_smtp:’ part, and add line

`1`	`interface = "${lookup{$sender_address_domain}lsearch{/etc/virtual/domainips}{$value}}"`

It should look something like:

1
2
3

remote_smtp:
  driver = smtp
  interface = "${lookup{$sender_address_domain}lsearch{/etc/virtual/domainips}{$value}}"

File /etc/virtual/domainips syntax:

`1`	`>domain.tld: ip.add.re.ss`

That’s it. Create the domainowners file, restart exim, and You are all set!

php5-fpm socket permission denied.

Tue, 21 Oct 2014 21:12:07 +0000

After updating the system, one error occured, despite that everything worked before.

2014/10/19 23:01:34 [crit] 27407#0: *11 connect() to unix:/tmp/php.sock failed (13: Permission denied) while connecting to upstream, client: 78.58.224.49, server: domain.tld, request: "GET / HTTP/1.1", upstream: "fastcgi://unix:/tmp/php.sock:", host: "domain.tld"

The reason of it is that despite the defined user:

1
2

user = www-data
group = www-data

The php startup script still created the socket with the owner of root, and nginx was not able to reach it:

changing mysql to innodb_file_per_table=1 with lvm and replication almost without downtime

Sun, 19 Oct 2014 20:23:17 +0000

This conversion was made with mysql 5.5, but it should work with older versions, as replications on mysql were introduced since mysql 3.23.15.

Don’t forget the [Disclaimer][1] !!!
It requires more than the basic knowledge in linux, lvm and mysql. I do NOT take ANY responsibility for Your data loss!!!

Long story short:

enable mysql bin logs
create lvm snapshot
mark master status
-
start 2'nd mysql instance from snapshot
dump all data
-
start new mysql instance with innodb_file_per_table=1 in new data_dir
start import of dumped data
start replication from master
-
when slave=master stop both servers
move slave data_dir to /var/lib/mysql and start main mysql instance.

LET’S BEGIN:

insmod: error inserting ‘/lib/xenblk.ko’: -1 Operation not permitted () when booting XEN VM

Tue, 19 Aug 2014 20:26:42 +0000

When getting error

Loading xenblk.ko module
ksign: module signed with unknown public key
- signature keyid: 99a6f9685d38f58d ver=3
insmod: error inserting '/lib/xenblk.ko': -1 Operation not permitted
Loading ehci-hcd.ko module
ksign: module signed with unknown public key
- signature keyid: 99a6f9685d38f58d ver=3
insmod: error inserting '/lib/ehci-hcd.ko': -1 Operation not permitted
Loading ohci-hcd.ko module
ksign: module signed with unknown public key
- signature keyid: 99a6f9685d38f58d ver=3
insmod: error inserting '/lib/ohci-hcd.ko': -1 Operation not permitted
Loading uhci-hcd.ko module
ksign: module signed with unknown public key
- signature keyid: 99a6f9685d38f58d ver=3
insmod: error inserting '/lib/uhci-hcd.ko': -1 Operation not permitted
Loading jbd.ko module
ksign: module signed with unknown public key
- signature keyid: 99a6f9685d38f58d ver=3
insmod: error inserting '/lib/jbd.ko': -1 Operation not permitted
Loading ext3.ko module
ksign: module signed with unknown public key
- signature keyid: 99a6f9685d38f58d ver=3
insmod: error inserting '/lib/ext3.ko': -1 Operation not permitted
Loading dm-mem-cache.ko module
ksign: module signed with unknown public key
- signature keyid: 99a6f9685d38f58d ver=3
insmod: error inserting '/lib/dm-mem-cache.ko': -1 Operation not permitted
Loading dm-mod.ko module
ksign: module signed with unknown public key
- signature keyid: 99a6f9685d38f58d ver=3
insmod: error inserting '/lib/dm-mod.ko': -1 Operation not permitted
Loading dm-log.ko module
ksign: module signed with unknown public key
- signature keyid: 99a6f9685d38f58d ver=3
insmod: error inserting '/lib/dm-log.ko': -1 Operation not permitted
Loading dm-region_hash.ko module
ksign: module signed with unknown public key
- signature keyid: 99a6f9685d38f58d ver=3
insmod: error inserting '/lib/dm-region_hash.ko': -1 Operation not permitted
Loading dm-message.ko module
ksign: module signed with unknown public key
- signature keyid: 99a6f9685d38f58d ver=3
insmod: error inserting '/lib/dm-message.ko': -1 Operation not permitted
Loading dm-raid45.ko module
ksign: module signed with unknown public key
- signature keyid: 99a6f9685d38f58d ver=3
insmod: error inserting '/lib/dm-raid45.ko': -1 Operation not permitted
Loading xennet.ko module
ksign: module signed with unknown public key
- signature keyid: 99a6f9685d38f58d ver=3
insmod: error inserting '/lib/xennet.ko': -1 Operation not permitted
Scanning and configuring dmraid supported devices
Creating root device.
Mounting root filesystem.
mount: could not find filesystem '/dev/root'
Setting up other filesystems.

Try to rebuild initrd for VM (it also forces resymlinking of kernel in use):

How-to start two skype instances on linux

Tue, 29 Jul 2014 15:55:46 +0000

Basically, for starting the second skype process, You need to have a separate skype data path.

So it can be started like that:

`1`	`skype --dbpath=~/.Skype2`

If You need all the history or settings of an existing skype, You may just copy that dir:

`1`	`cp -pr ~/.Skype ~/.Skype2`

p.s. You may start even more instances in the same way.

client denied by server configuration: owncloud and directadmin

Mon, 24 Mar 2014 19:38:19 +0000

If got error with owncloud when using owncloud application (custombuild managed structure):

1
2

[Mon Mar 24 19:42:39.379283 2014] [access_compat:error] [pid 525512] [client 86.100.237.21:55665] AH01797: client denied by server configuration: /home/admin/domains/domain.tld/
private_html/owncloud/remote.php

Solution is to enable PUT requests.

need to edit:
/etc/httpd/conf/extra/httpd-directories.conf
and add PUT to these lines:

<Limit GET POST OPTIONS PROPFIND PUT>
                Order allow,deny
                Allow from all
        </Limit>
        <LimitExcept GET POST OPTIONS PROPFIND PUT>
                Order deny,allow
                Deny from all
        </LimitExcept>

Don’t forget to copy httpd-directories.conf to custom directory, as it will be rewrote after doing updates with directadmin!!!

Power OFF and ON USB device in linux (ubuntu)

Thu, 20 Feb 2014 11:43:32 +0000

There are two methods of doing this. So first method is for kernels after 2.6.32, and second for older ones.

To view kernel version, simply enter:

1
2

> uname -r 
3.8.0-29-generic

find Your USB device, no mather what it is USB flash drive, Yubikey, USB disk drive:

> lsusb 
Bus 001 Device 002: ID 8087:0024 Intel Corp. Integrated Rate Matching Hub
Bus 002 Device 002: ID 8087:0024 Intel Corp. Integrated Rate Matching Hub
Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
Bus 002 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
Bus 002 Device 005: ID 12d1:14db Huawei Technologies Co., Ltd. 
Bus 002 Device 006: ID 0424:2514 Standard Microsystems Corp. USB 2.0 Hub

> lsusb -t
/:  Bus 02.Port 1: Dev 1, Class=root_hub, Driver=ehci-pci/2p, 480M
    |__ Port 1: Dev 2, If 0, Class=hub, Driver=hub/6p, 480M
        |__ Port 1: Dev 5, If 0, Class=comm., Driver=cdc_ether, 480M
        |__ Port 1: Dev 5, If 1, Class=data, Driver=cdc_ether, 480M
        |__ Port 5: Dev 6, If 0, Class=hub, Driver=hub/4p, 480M
/:  Bus 01.Port 1: Dev 1, Class=root_hub, Driver=ehci-pci/2p, 480M
    |__ Port 1: Dev 2, If 0, Class=hub, Driver=hub/6p, 480M

Our target is:

Squirrelmail hangs forever (DNS cache problem)

Tue, 18 Feb 2014 12:04:10 +0000

Noticed that one squirrelmail sometime hangs.
It contained default directadmin installation configs.

Straced proccess, and found that it queries DNS servers for a quite long time (only snippet):

...
munmap(0xb7ef5000, 4096)                = 0
socket(PF_INET, SOCK_DGRAM, IPPROTO_IP) = 2108
connect(2108, {sa_family=AF_INET, sin_port=htons(53), sin_addr=inet_addr("127.0.0.1")}, 28) = 0
fcntl64(2108, F_GETFL)                  = 0x2 (flags O_RDWR)
fcntl64(2108, F_SETFL, O_RDWR|O_NONBLOCK) = 0
gettimeofday({1392720955, 282214}, NULL) = 0
poll([{fd=2108, events=POLLOUT}], 1, 0) = 1 ([{fd=2108, revents=POLLOUT}])
send(2108, "H\177\1\0\0\1\0\0\0\0\0\0\00248\003177\003231\003173\tspam"..., 54, MSG_NOSIGNAL) = 54
poll([{fd=2108, events=POLLIN}], 1, 5000) = 1 ([{fd=2108, revents=POLLIN}])
ioctl(2108, FIONREAD, [104])            = 0
recvfrom(2108, "H\177\201\203\0\1\0\0\0\1\0\0\00248\003177\003231\003173\tspam"..., 1024, 0, {sa_family=AF_INET, sin_port=htons(53), sin_addr=inet_addr("127.0.0.1")}, [16]) = 104
close(2108)                             = 0
time(NULL)                              = 1392720955
open("/etc/hosts", O_RDONLY)            = 2108
fcntl64(2108, F_GETFD)                  = 0
fcntl64(2108, F_SETFD, FD_CLOEXEC)      = 0
fstat64(2108, {st_mode=S_IFREG|0644, st_size=273, ...}) = 0
mmap2(NULL, 4096, PROT_READ|PROT_WRITE, MAP_PRIVATE|MAP_ANONYMOUS, -1, 0) = 0xb7ef5000
read(2108, "# Do not remove the following li"..., 4096) = 273
read(2108, "", 4096)                    = 0
close(2108)                             = 0
munmap(0xb7ef5000, 4096)                = 0
socket(PF_INET, SOCK_DGRAM, IPPROTO_IP) = 2108
connect(2108, {sa_family=AF_INET, sin_port=htons(53), sin_addr=inet_addr("127.0.0.1")}, 28) = 0
fcntl64(2108, F_GETFL)                  = 0x2 (flags O_RDWR)
fcntl64(2108, F_SETFL, O_RDWR|O_NONBLOCK) = 0
gettimeofday({1392720955, 396610}, NULL) = 0
poll([{fd=2108, events=POLLOUT}], 1, 0) = 1 ([{fd=2108, revents=POLLOUT}])
send(2108, "\242Z\1\0\0\1\0\0\0\0\0\0\00248\003177\003231\003173\5dnsb"..., 48, MSG_NOSIGNAL) = 48
poll([{fd=2108, events=POLLIN}], 1, 5000) = 0 (Timeout)
socket(PF_INET, SOCK_DGRAM, IPPROTO_IP) = 2109
connect(2109, {sa_family=AF_INET, sin_port=htons(53), sin_addr=inet_addr("8.8.8.8")}, 28) = 0
fcntl64(2109, F_GETFL)                  = 0x2 (flags O_RDWR)
fcntl64(2109, F_SETFL, O_RDWR|O_NONBLOCK) = 0
gettimeofday({1392720960, 396031}, NULL) = 0
poll([{fd=2109, events=POLLOUT}], 1, 0) = 1 ([{fd=2109, revents=POLLOUT}])
send(2109, "\242Z\1\0\0\1\0\0\0\0\0\0\00248\003177\003231\003173\5dnsb"..., 48, MSG_NOSIGNAL) = 48
poll([{fd=2109, events=POLLIN}], 1, 3000) = 0 (Timeout)
socket(PF_INET, SOCK_DGRAM, IPPROTO_IP) = 2110
connect(2110, {sa_family=AF_INET, sin_port=htons(53), sin_addr=inet_addr("8.8.4.4")}, 28) = 0
fcntl64(2110, F_GETFL)                  = 0x2 (flags O_RDWR)
fcntl64(2110, F_SETFL, O_RDWR|O_NONBLOCK) = 0
gettimeofday({1392720963, 399893}, NULL) = 0
poll([{fd=2110, events=POLLOUT}], 1, 0) = 1 ([{fd=2110, revents=POLLOUT}])
send(2110, "\242Z\1\0\0\1\0\0\0\0\0\0\00248\003177\003231\003173\5dnsb"..., 48, MSG_NOSIGNAL) = 48
poll([{fd=2110, events=POLLIN}], 1, 6000) = 1 ([{fd=2110, revents=POLLIN}])
ioctl(2110, FIONREAD, [48])             = 0
recvfrom(2110, "\242Z\201\202\0\1\0\0\0\0\0\0\00248\003177\003231\003173\5dnsb"..., 1024, 0, {sa_family=AF_INET, sin_port=htons(53), sin_addr=inet_addr("8.8.4.4")}, [16]) = 48
close(2108)                             = 0
close(2109)                             = 0
close(2110)                             = 0
socket(PF_INET, SOCK_DGRAM, IPPROTO_IP) = 2108
connect(2108, {sa_family=AF_INET, sin_port=htons(53), sin_addr=inet_addr("127.0.0.1")}, 28) = 0
fcntl64(2108, F_GETFL)                  = 0x2 (flags O_RDWR)
fcntl64(2108, F_SETFL, O_RDWR|O_NONBLOCK) = 0
gettimeofday({1392720968, 243445}, NULL) = 0
poll([{fd=2108, events=POLLOUT}], 1, 0) = 1 ([{fd=2108, revents=POLLOUT}])
send(2108, "\242Z\1\0\0\1\0\0\0\0\0\0\00248\003177\003231\003173\5dnsb"..., 48, MSG_NOSIGNAL) = 48
poll([{fd=2108, events=POLLIN}], 1, 5000) = 0 (Timeout)
socket(PF_INET, SOCK_DGRAM, IPPROTO_IP) = 2109
connect(2109, {sa_family=AF_INET, sin_port=htons(53), sin_addr=inet_addr("8.8.8.8")}, 28) = 0
fcntl64(2109, F_GETFL)                  = 0x2 (flags O_RDWR)
fcntl64(2109, F_SETFL, O_RDWR|O_NONBLOCK) = 0
gettimeofday({1392720973, 266865}, NULL) = 0
poll([{fd=2109, events=POLLOUT}], 1, 0) = 1 ([{fd=2109, revents=POLLOUT}])
send(2109, "\242Z\1\0\0\1\0\0\0\0\0\0\00248\003177\003231\003173\5dnsb"..., 48, MSG_NOSIGNAL) = 48
poll([{fd=2109, events=POLLIN}], 1, 3000) = 0 (Timeout)
socket(PF_INET, SOCK_DGRAM, IPPROTO_IP) = 2110
connect(2110, {sa_family=AF_INET, sin_port=htons(53), sin_addr=inet_addr("8.8.4.4")}, 28) = 0
fcntl64(2110, F_GETFL)                  = 0x2 (flags O_RDWR)
fcntl64(2110, F_SETFL, O_RDWR|O_NONBLOCK) = 0
gettimeofday({1392720976, 278949}, NULL) = 0
poll([{fd=2110, events=POLLOUT}], 1, 0) = 1 ([{fd=2110, revents=POLLOUT}])
send(2110, "\242Z\1\0\0\1\0\0\0\0\0\0\00248\003177\003231\003173\5dnsb"..., 48, MSG_NOSIGNAL) = 48
poll([{fd=2110, events=POLLIN}], 1, 6000) = 1 ([{fd=2110, revents=POLLIN}])
ioctl(2110, FIONREAD, [48])             = 0
recvfrom(2110, "\242Z\201\202\0\1\0\0\0\0\0\0\00248\003177\003231\003173\5dnsb"..., 1024, 0, {sa_family=AF_INET, sin_port=htons(53), sin_addr=inet_addr("8.8.4.4")}, [16]) = 48

...

After ‘querying’ for some time, it generated file:

MX backup server with postfix on debian

Sun, 16 Feb 2014 14:13:06 +0000

Not to lose mails when main mail server is down, best solution is - mx backup server.

This is configuration to have backup mx server set up and running.

myhostname = mx2.loginroot.com
mynetworks = 127.0.0.0/8
maximal_queue_lifetime = 120d
smtpd_banner = $myhostname ESMTP

relay_domains = hash:/etc/postfix/relaydomains
transport_maps = hash:/etc/postfix/transportmaps

smtpd_recipient_restrictions =
 permit_mynetworks,
 reject_unauth_destination

create /etc/postfix/relaydomains with contents:

RC car steering solution?

Thu, 13 Feb 2014 07:23:13 +0000

One solution came in mind for RC car steering.
Control servo (planned or RC car steering) with potentiometer.

Testing setup is very simple.
Servo setup: +, - and control pin. Same as here:
Simple servo setup

Potentiometer will work as voltage divider.
Here is more info about it: Voltage divider.

Connection to arduino would be also pretty much simple:

Potentiometer is like this whole divider system. So when potentiometer is with big enough resistivity (I think mine is 10Kohms) - One end goes to 5v, other to ground, and the one in the middle, goes to analog read.
In this setup voltage output goes from 0V (when top resistor is 0ohms) to 5V (when bottom resistor is 0ohms).

New project idea – RC car

Sun, 09 Feb 2014 12:36:17 +0000

One day thought that would be nice to make a RC car controlled with arduino.
Unfortunately I don’t know how make such hardware, I needed to to find any car, that has possibility to turn front wheels to both sides.
After few days of searching (all of them were expensive or front wheels were ‘fixed’ they were able only to rotate clock or co-clock wise) found this:

As it was broken and unpacked, it cost only 13$. It was even cheaper than cars without any RC stuff (anyway it will be thrown out).

Arduino + servo simple setup

Wed, 05 Feb 2014 13:32:41 +0000

Using servo with arduino is pretty easy.

Here is my setup:

Servo has 3wires, vcc, gnd and control.
My servo colors were
+5V - red,
gnd - brown
control - orange.

Code for arduino (my control wire goes to pin12):

Servo myServo;

int servoPin = 12;

void setup() {
	myServo.attach(servoPin);
}

void loop() {
	myServo.write(0);
	delay(1000);
	myServo.write(180);
	delay(1000);
}

#include - we load servo library
Servo myServo; - create an servo object
With int servoPin we just make simple variable to be more readable in code (currently doesn’t make sense as there is only one pin in use, but it’s a good habit to have).
myServo.attach(servoPin); - point to servo object to use ‘servoPin’ pin.
myServo.write() function takes degrees as parameters.
My current servo takes from 0 to 180.

Exim mail via different IP

Tue, 04 Feb 2014 16:03:16 +0000

In /etc/exim.conf
You need to change:

1
2

remote_smtp:
  driver = smtp

1
2
3

remote_smtp:
  driver = smtp
  interface = 11.22.33.44

CP2102 + Arduino pro mini HOW TO

Sun, 02 Feb 2014 14:30:47 +0000

One day I noticed, that arduino pro mini are quite cheap for everyday stuff.
It’s very small, and to keep lower price, this Arduino does not have USB interface.
On ebay it’s ~3$:

For writing sketches to this Arduino, bought CP2102 USB to UART brige:

It’s easy to connect. Arduino has few pins with the same name, so You may choose any of them. I personally didn’t even solder that additional part of arduino mini pro in front, as same pins were on side anyway.

New toy arrived!!!

Mon, 27 Jan 2014 20:19:34 +0000

Recently, a new toy arrived:

So it’s a good opportunity to create new category

As is written in [Disclaimer][2] - I do not take any complaints if You brake anything following my blog.
It doesn’t matter if You order arduino or funduino (china clone of arduino), it’s opensource project, and You can still use development tools from official arduino.cc site.
Arduino IDE is here.
So first task is to check if it works :)

Openvpn centos 6.4 x64 with certificates

Fri, 29 Nov 2013 17:44:09 +0000

**Renewed POST with new info, as openvpn package changed, not everything in old tutorial is true now :)

This tutorial is based by old one

Default centos repositoryt doesn’t have openvpn package, so lets add epel repository first.

I downloaded it from this mirror here.

`1`	`yum install http://dl.fedoraproject.org/pub/epel/6/x86_64/epel-release-6-8.noarch.rp`

(link may be different depending on latest release version, currently file is epel-release-6-8.noarch.rpm)

After that, You can install openvpn and easy-rsa packages:

bandwidth monitoring script for nagios (CentOS) – bw_watch.sh

Mon, 14 Oct 2013 14:18:12 +0000

Bash script for nagios, that monitors bandwidth usage on remote host.
# usage ./bw_watch bw_warning bw_critical pkt_warning pkt_critical
It takes momentary 5 secs average bandwidth from ‘vnstat’, and returns value for nagios.

bc and vnstat packages at host side are necessary.
bc package should be in default centos repository, vnstat and nrpe You can find in EPEL repo.

Script itself, that You have to put to nagios plugins directory. In my case it was: /usr/lib64/nagios/plugins/

Horde5 webmail: sending mail via smtp.

Mon, 15 Jul 2013 08:06:16 +0000

You can fix error

`1`	`There was an error sending your message: sendmail: unknown error (127)`

by changing sending method to SMTP:

You have to login to webmail using administration webmail (it was entered during horde5 instalation).
Or You may enter later in file $horde_docroot$/config/conf.php
in lin

`1`	`$conf\['auth'\]\['admins'\] = array('enter@your\_existing\_mail.tld');`

After logging in with admin credentials choose on top menu gear icon, then ‘Horde (horde)’ -> Mailer
And change ‘$conf[mailer][type]’ to SMTP server.
Also enter these lines

evolution-ews plugin install script for evolution 3.2.3 on ubuntu 12.04

Mon, 12 Nov 2012 09:26:18 +0000

Made script for automatic evolution-ews plugin install.
By default ubuntu 12.04 uses evolution 3.2.3 and there are no evolution-ews package in repository.
Use evolution-ews plugin the same version as Your evolution mail client version.

If You want to compile and edit configuration file manualy, script automates everything that is written in this post:
Adding MS exchange compability with evolution-ews in ubuntu 12.04

Download link:
evolution-ews-3.2.3-install.sh

usage:

1
2
3

wget https://loginroot.com/scripts/evolution-ews-3.2.3-install.sh
chmod +x evolution-ews-3.2.3-install.sh
./evolution-ews-3.2.3-install.sh

That’s it!!!

Adding MS exchange compability with evolution-ews in ubuntu 12.04

Sun, 11 Nov 2012 12:48:15 +0000

Ubuntu 12.04 has crappy support for bussiness solutions.
You can check MS exchange mail, synchronize calendar in evolution with evolution-ews which is missing in default repository. Exchange has to have owa enabled.

Also default client on Ubuntu precise is thunderbird, which doesn’t have ability to use system provided calendar.
How silly is that?

Didn’t want to add external repositories, as it updates lots of packaged in system, and crappy apt-get doesn’t do its job to keep only one version!!! I needed only one small plugin, so why should I touch all default system, which suppose to be ‘stable’?

Ubuntu 12.04 Wireless unmanaged

Fri, 09 Nov 2012 20:10:19 +0000

When installing Ubuntu, and during install You choose to use wlan0 interface.
On first system boot wireless becomes unmanaged, and You can’t choose wireless AP.

solution is to edit file /etc/network/interfaces
There should be lines like

auto wlan0
iface wlan0 inet dhcp
   wpa-ssid wifi-name
   wpa-psk wifi-key

delete those lines and reboot computer. Network manager will be able to control wifi.

how to: create openvpn centos 6.3 x64 with certificates

Wed, 03 Oct 2012 11:59:34 +0000

Note: Some parts of this tutorial are not working anymore. New version available here Default centos repositoryt doesn’t have openvpn package, so lets add epel repository first.

I downloaded it from this mirror here. downloading repo:

`1`	`# wget http://mirror.duomenucentras.lt/epel/6/x86_64/epel-release-6-7.noarch.rp`

and installing repo:

`1`	`# yum localinstall epel-release-6-7.noarch.rp`

After that, You can install openvpn package:

`1`	`# yum install openvpn.x86_6`

cd to easy-rsa subdirectory in openvpn catalog.

reset network interfaces on centos 6.3

Tue, 02 Oct 2012 13:17:42 +0000

If You copy/clone Your machine, or change network card, on linux Your interfaces get names eth1, eth2, eth3 and so on.

Before makeing changes just delete /etc/udev/rules.d/70-persistent-net.rules file. After first boot, network interfaces will be enumerated from zero again.

`1`	`rm /etc/udev/rules.d/70-persistent-net.rules`

tune2fs, make free space to appear

Fri, 21 Sep 2012 17:20:48 +0000

ext2 ant ext3 file systems has reserved 5% disk space in case of dissapeared free space, for system proccesses. That system wouldn’t crash and would keep going.
For big filesystems those 5% are too much(with 1TB it would be 50G for system procceses), so we can tune a little bit.
You can check how much reserved space partition has with (You need root privileges for all actions):

$ sudo tune2fs -l /dev/sda1 | grep Reserved 
Reserved block count:     12441
Reserved GDT blocks:      256
Reserved blocks uid:      0 (user root)
Reserved blocks gid:      0 (group root)

to change reserved space (in percents)

locale.Error: unsupported locale (Solved)

Wed, 13 Jun 2012 19:31:32 +0000

# iotop
Traceback (most recent call last):
  File "/usr/bin/iotop", line 16, in <module>
    main()
  File "/usr/lib/pymodules/python2.6/iotop/ui.py", line 506, in main
    locale.setlocale(locale.LC_ALL, '')
  File "/usr/lib/python2.6/locale.py", line 513, in setlocale
    return _setlocale(category, locale)
locale.Error: unsupported locale setting

Solution:

apache [warn] module ssl_module is already loaded, skipping (Solved)

Sat, 09 Jun 2012 13:13:02 +0000

1
2
3

# httpd -t
[Fri Jun 08 11:12:19 2012] [warn] module ssl_module is already loaded, skipping
Syntax OK

This happens, because ssl module is enabled in more than 1 place:

1
2
3

# grep ssl_module -rI /etc/httpd/*
conf/httpd.conf:LoadModule ssl_module /usr/lib64/httpd/modules/mod_ssl.so
conf.d/ssl.conf:LoadModule ssl_module modules/mod_ssl.so

we load all needed modules in one place - /etc/httpd/conf/httpd.conf.
So we can safely comment out in /etc/httpd/conf.d/ssl.conf file.

After upgrade: PHP Warning: PHP Startup: Unable to load dynamic library (Solved)

Fri, 08 Jun 2012 08:09:02 +0000

After upgrade to php53

#  php -v
PHP Warning:  PHP Startup: Unable to load dynamic library '/usr/lib64/php/modules/dbase.so' - /usr/lib64/php/modules/dbase.so: cannot open shared object file: No such file or directory in Unknown on line 0
PHP Warning:  PHP Startup: Unable to load dynamic library '/usr/lib64/php/modules/ncurses.so' - /usr/lib64/php/modules/ncurses.so: undefined symbol: first_arg_force_ref in Unknown on line 0
PHP 5.3.13 (cli) (built: May  9 2012 09:51:06) 
Copyright (c) 1997-2012 The PHP Group
Zend Engine v2.3.0, Copyright (c) 1998-2012 Zend Technologies

solution is to go to ‘/etc/php.d’, and disable those .ini files:

Wrapper /var/www/cgi-bin/cgi_wrapper/cgi_wrapper cannot be accessed: (2)No such file or directory

Sun, 06 May 2012 09:10:38 +0000

If You got messed up with PHP in CentOS with plesk 10.4, You might get such error:

1
2
3

# httpd -t
Syntax error on line 65 of /var/www/vhosts/domain.tld/conf/13353493930.96968200_httpd.include:
Wrapper /var/www/cgi-bin/cgi_wrapper/cgi_wrapper cannot be accessed: (2)No such file or directory

Small fix to make it work:

`1`	`# ln -s /usr/bin/sw-engine-cgi /var/www/cgi-bin/cgi_wrapper/cgi_wrapper`

sw-engine-cgi is from PSA repository:

1
2

# yum list | grep sw-engine
sw-engine.x86_64                       2.3-201111031522.centos6      @PSA_10_4_4-dist

Well, You can use different cgi wrapper.
In my system there were 3 present.

Ubuntu 12.04 64bit /sbin/insserv: No such file or directory

Fri, 04 May 2012 19:31:40 +0000

Small fix if You have an error when You try to turn on startup programs.

for example:

1
2

# chkconfig -s ssh on
/sbin/insserv: No such file or directory

Fix is simple, just to add symlink:

`1`	`# ln -s /usr/lib/insserv/insserv /sbin/insserv`

[paypal-donation]

Creating debian template for lxc virtualization

Mon, 20 Feb 2012 17:33:44 +0000

To create debian template for lxc you need fresh copy of debian system.

Well, we do have one installed recently on guruplug.
All debian install on guruplug steps are here.

So, lets archive all system to tar.gz.

Firstly it’s a good practice to exclude unneeded files.
Create /tmp/exclude file with this content:

# cat /tmp/exclude 
/tmp/*
/proc/*
/dev/*
/sys/*
/mnt/*
/usr/src/*

Edit templates /etc/fstab file and comment everything out.
Create new file /etc/fstab.lxc
Contents of that file:

Adding bridge interface for lxc containers

Sat, 18 Feb 2012 11:54:18 +0000

You can make a bridge by changing network configuration file:

# cat /etc/network/interfaces 
# This file describes the network interfaces available on your system
# and how to activate them. For more information, see interfaces(5).

# The loopback network interface
auto lo
iface lo inet loopback

# The primary network interface
allow-hotplug eth0
# iface eth0 inet dhcp

# make this manual, that network manager would not conflict when you connect utp cable
iface eth0 inet manual

auto br0
iface br0 inet static
	bridge_ports eth0
	address 192.168.1.10
	broadcast 192.168.1.255
	netmask 255.255.255.0
	gateway 192.168.1.1

In this configuration I used br0 static configuration.
If You need dhcp, You can use this configuration:

Could not chdir to home directory /home/nsc: Permission denied /bin/bash: Permission denied Connection to 192.168.1.13 closed.

Mon, 13 Feb 2012 08:14:25 +0000

after creating lxc container, I wasn’t able to connect to it as a simple user. Also couldn’t change to simple user using ‘su’.

All I got was an error:

1
2
3

Could not chdir to home directory /home/nsc: Permission denied
/bin/bash: Permission denied
Connection to 192.168.1.13 closed.

All permissions of home directory were correct, as to be 100% sure, deleted user and created again. Still the same problem.

sudo without password

Sun, 12 Feb 2012 13:27:18 +0000

This line should do the job:

`1`	`your_username ALL=(ALL) NOPASSWD: ALL`

You can edit sudoers file by entering command ‘visudo’.

varnish+nginx as backend and real IP

Sun, 12 Feb 2012 12:50:13 +0000

With default config nginx all IP addresses shows as localhost.
Here is configuration to show real client IP.

Snippet from varnish config:

sub vcl_pipe {
        set bereq.http.connection = "close";
        if (req.http.X-Forwarded-For) {
                set bereq.http.X-Forwarded-For = req.http.X-Forwarded-For;
        } else {
                set bereq.http.X-Forwarded-For = regsub(client.ip, ":.*", "");
          }
    }
    sub vcl_pass {
        set bereq.http.connection = "close";
        if (req.http.X-Forwarded-For) {
                set bereq.http.X-Forwarded-For = req.http.X-Forwarded-For;
        } else {
          set bereq.http.X-Forwarded-For = regsub(client.ip, ":.*", "");
          }
    }

Snippet from nginx config:

Replace string in all files using sed in linux

Fri, 10 Feb 2012 18:08:03 +0000

Replaces all ‘OLDSTRING’ to ‘NEWSTRING’ in all files

`1`	`$ grep -rl OLDSTRING * \| sort \| uniq \| xargs sed -i -e ‘s/OLDSTRING/NEWSTRING/’`

or using find:

`1`	`find /public_html/ -iname '.php' -exec sed -i -e 's/OLDSTRING/NEWSTRING/' {} \;`

Installing debian to guruplug SD card

Mon, 30 Jan 2012 21:43:00 +0000

Before installing debian, You need to prepare tftp server and connect to guruplug via serial console.

Also You’ll need UTP cable for network connection and SD card for installing the base system.

Lets prepare debian install.
Download uImage and uInitrd images from debian site:
ftp://ftp.debian.org/debian/dists/stable/main/installer-armel/current/images/kirkwood/netboot/marvell/sheevaplug/

So, put those files to tftp server directory (on your pc).
I put to /tftproot/install/* directory.

Put SD card flash to guruplug, connect UTP cable and connect to it via serial cable.

Centos wordpress php-fpm + nginx + apc

Sat, 21 Jan 2012 09:47:32 +0000

If you tired of having same old apache, and want something more fast and reliable, you can install nginx + php-fpm.

Firstly you need to install new repo, to have ability to add those packages. And php-fpm works out of the box (without need to patch PHP) with version PHP 5.3. Official repository does not have this version yet.

`1`	`wget -q -O - http://www.atomicorp.com/installers/atomic.sh \| sh`

`1`	`yum install nginx php-fpm php-pecl-apc`

When you have more virtual hosts than one, and you want to have more convenient way to manage them, well to turn them on and off if being more precisely.
You can create two config directories. One directory ‘sites-available’ to gave all nginx virtual host configs, and ‘sites-enabled’, which will be included in main nginx config file, and all symbolic links would be kept inside this direcotory pointing to real config files.

Installing uboot to guruplug server plus

Thu, 19 Jan 2012 19:52:54 +0000

The first thing you have to do, is to read this:
[https://loginroot.com/disclaimer/][1]

Before installing uboot to guruplug server plus you have to set few things.

You have to be able to connect to Guruplug via serial connection.
The second thing, is to have tftp server up and running.

Also you need to have utp cable (to transfer the image to guruplug) and the boot image itself.

I used this image for my guruplug:
http://people.debian.org/~tbm/u-boot/2011.12-2/guruplug/u-boot.kwb

Disclaimer

Thu, 19 Jan 2012 19:31:00 +0000

All the content written on this blog is for informational use only.

The steps taken in the guides are not necessary the recommended ones, due to the fact that not all the tasks align with a personal mindset. The blog/diary is more like a historical reference or a login case the subject needs to be revisited in the future.

I DO NOT take any responsibility for actions taken on the basis of this blog.

Guruplug serial connection

Tue, 17 Jan 2012 21:11:06 +0000

If you cannot connect to guruplug for some reason via lan interface, for example sshd stopped running or just firewalled yourself out of it, you can connect to it using serial connection.

When making order guruplug from original site, by adding few more bucks you can get development kit.

For serial connection you need only smaller connector, the bigger one is used for jtag connection, which is used if you brick the device.

Create tftp server on ubuntu 11.10

Tue, 17 Jan 2012 19:37:07 +0000

For some tasks, like loading bootloader you need to have tftp server.
Here are few steps to set it up and get it running.

The easiest way to get tftp is from repository.
So, lets start:

`1`	`$ sudo apt-get install tftpd`

After install finished you have to create tftp configuration file:

`1`	`$ sudo vim /etc/xinetd.d/tftp`

Press letter ‘i’ for insert mode end paste these lines:

Recover mysql root password

Sun, 15 Jan 2012 17:32:40 +0000

It happens quite often, that You need to do something on database, which was installed long long time ago. And nobody remembers the root password.

Kinda good practice is to save mysql password in root user directory with a ‘.’ in front of it (hidden), for example .my.cnf Nobody can read root users directory, except root, and anyway, root user can change mysql password at any time, so it’s not an security issue.
If all users are able to see root users directory, that is a problem.

Hello world!

Sun, 15 Jan 2012 15:47:14 +0000

Today, new blog is born.

`1`	`print "hello world!\nThis is first post";`