SIP over TCP in Asterisk

Do sip.conf treba:

tcpenable=yes ; Enable server for incoming TCP connections (default is no)
tcpbindaddr=0.0.0.0 ; IP address for TCP server to bind to (0.0.0.0 binds to all interfaces)

Do users.conf pre jednotliveho uzivatela, pripadne globalne do sip.conf treba:

[6000]
...
transport=tcp
...

iLBC & SPEEX kodek na Asterisku

Kodeky s ktorymi som experimentoval hlavne koli potrebe co najmensieho datoveho toku pri telefonovani cez 3G siete.

Instalacia iLBC kodeku

Od 1.6 verzie Asterisku nie je iLbc standartnou vybavou, treba stiahnut rucne.. Asterisk ma na to vsak pripraveny script, kde po odsuhlaseni licencie stiahne vsetko potrebne:

cd /usr/src/asterisk-*/
./contrib/scripts/get_ilbc_source.sh
./configure
make menuconfig (vybrat iLbc v codecs sekcii)
make install

Nasledne treba povolit codec bud globalne v sip.conf, alebo pre kazdeho uzivatela zvlast:

disallow=all
allow=ilbc
allow=dalsikodek

A nakoniec kontrola:

asterisk -vvvvvvvvvvvvvr
load codec_ilbc.so
core show codecs
core show translations

Odkazy pre iLBC:

http://www.voip-info.org/wiki/view/iLBC

http://blogs.elastix.org/en/2009/11/ilbc-vs-g729-the-quick-guide-to-using-compressed-codecs-in-elastix/

Instalacia SPEEX kodeku
Na Debiane napiseme nasledovne:

aptitude install speex libspeex-dev

Potom je potrebne skompilovat Asterisk. Najprv spustime ./configure a nasledne v menuconfig oznacime podporu kodeku SPEEX v sekcii Codec Translators. Nasleduje obligatne make a make install.

V pripade,ze uz mame Asterisk skompilovany a kodek chceme pridat, musime spustit make distclean, aby sme Asteriska prinutili si podporu SPEEX natiahnut.

V pripade potreby rucnej kompilacie kodeku, je potrebne stiahnut si zdrojaky z adresy:
http://downloads.us.xiph.org/releases/speex/

How to Boinc on Linux (Debian)

At first:

aptitude install boinc-client boinc-manager

Then you have to create password for localhost, so open file and write your password on it.:

vim /etc/boinc-client/gui_rpc_auth.cfg

Restart client:

/etc/init.d/boinc-client restart

Next we want to attach to some Boinc project.

Find out your account key:

boinc_cmd --host localhost --passwd your_localhost_password --lookup_account http://einstein.phys.uwm.edu/ account_name account_password

Connect to project with your account key:

boinc_cmd --host localhost --passwd your_localhost_password --project_attach http://einstein.phys.uwm.edu/ your_account_key

Check Boinc state:

boinc_cmd --host localhost --passwd your_localhost_password --get_state

P.S.
Boinc client and his tasks usually running with nice 19.

Remove service after start on Debian/*buntu

update-rc.d -f service remove

CentOS a divne LVM/CentOS installation strange LVM partition

SK 

Pokial sa snazite nainstalovat CentOS a nebodaj spravit softwarovy RAID medzi dvoma diskami a v rozdelovaci diskov vidite miesto dvoch diskov len jednu divnu LVM particiu s ktorou sa neda pohnut tak vedzte, ze pri boote musite do konzole pripisat nasledovne:

linux nodmraid

A kedze CentOS nema vlastny zoznam mirrorov a musite rucne vypisovat odkial si ma netinstall stiahnut potrebny image, tak tu je jeden ktory sa zda byt celkom rychly:

http://mirror.karneval.cz/pub/centos/5.3/os/x86_64

do koloniek musite vypisat najskor mirror.karneval.cz a do druhej zvysok cize /pub/centos/5.3/os/x86_64.  

ENG

If you want install CentOS plus configure software RAID and you still see in partition manager only one strange LVM partition without any posibbility to change it and your hair is standing up (like my was), continue reading. In boot console write this:

linux nodmraid

And done :) CentOS doesn`t have preset mirror (or posibility to choose one), you must write an URL to installator. In Czech republic I`m using http://mirror.karneval.cz/pub/centos/5.3/os/x86_64. In first textbox write mirror.karneval.cz, in second /pub/centos/5.3/os/x86_64. I don`t understand that policy, because this is a huge security problem, but maybe there is some reason :) Fortunately I`m using Debian.

Ako vytvorit /dev/null ?/How to create /dev/null ?

SK

Takto:

mknod -m 0666 /dev/null c 1 3

ENG

Here is the command:

mknod -m 0666 /dev/null c 1 3

Ako obnovit Primary/Primary DRBD replikaciu?

Pri pade jedneho zo servrov ktorych niektora particia bezi na Primary/Primary DRBD je potrebne odsranit problem so Split Brain-om, ktory je dobre viditelny aj v logu (Split-Brain detected, dropping connection!). Teoria je taka, ze DRBD sa nevie rozhodnut, ktory z nodov ma aktualne data a preto je treba mu to povedat rucne. Ako na to?

Na xene ktory spadol a ma teda starsie data pouzijeme tento subor prikazov:

drbdadm secondary r0
drbdadm -- --discard-my-data connect r0

• V prvom kroku urcime ze ze nadalej sa ma node spravat ako secondary, teda spravime Primary/Secondary replikaciu
• V druhom kroku mu povieme ze jeho data su stare a teda ma brat ohlad na data, ktore su na druhom node.
• r0 je resource definovany v konfiguraku, ktory sa nachadza v /etc/drbd.conf

Na xene, ktory nespadol a teda ma aktualne data pouzijeme tento subor prikazov:

drbdadm connect r0
drbdadm -- -o primary all

• V prvom kroku povieme aby sa nody znovu spojili, kedze druhy nod ma priznak ze je neaktualny, synchronizuju sa na neho data z aktualneho nodu.
• Druhy prikaz je mozne zadat na hociktorom z nodov, znamena ze z oboch nodov spravi opat Primary.

Pokial nenastartovala OCFS2 vrstva, spravime tak klasickym /etc/init.d/ocfs2 start . OCFS2 si aj automaticky namountuje particiu. A to je vsetko :)