gomez’s blog

Obavještavamo sve zainteresovane da je objavljena finalna verzija Bosanskohercegovačkog Linux desktopa BHLD 2.0. Ovaj projekat je finansiralo Ministarstvo obrazovanja, nauke i kulture FBiH, a vodio ga je dr Samir Ribić sa ETFa Sarajevo. U projektu su učestvovali i mnogobrojni studenti ETFa (kojima se ovom prilikom zahvaljujemo).

Bh. Linux desktop je projekat čiji je cilj prezentovanje Linuxa u svojstvu radne stanice (desktopa) kao i aktivnosti ULK na lokalizaciji (prevođenu na naše jezike) grafičkog okruženja i aplikacija. BHLD 2.0 dolazi sa setom aplikacija posebno prilagođenim za primjenu u obrazovanju, kao i sa vrlo detaljnim priručnikom na našem jeziku u kojem su opisane sve aplikacije uključene u distribuciju. Od aplikacija ističemo OpenOffice.org 3.1 i Firefox 3.0. Pored standardnog KDE okruženja BHLD 2.0 sadrži i lagano okruženje za starije računare te bi trebao biti upotrebljiv i na računarima sa svega 128 MB RAM. BHLD je “live CD” što znači da ga možete isprobati bez instalacije, a za razliku od verzije 1.x instalacija na disk je potpuno podržana.

Verzija 2.0 bazirana je na megapopularnoj Ubuntu Linux distribuciji. Tehničke detalje (spisak paketa) možete saznati na ULK wiki stranici koja će se po potrebi dopunjavati informacijama:
http://wiki.linux.org.ba/BHLD2

BHLD 2.0 moći ćete naći u sljedećem broju magazina “Info” zajedno sa člankom i intervjuom, a ISO image možete preuzeti sa mirror servera:
ftp://mirror.bhld.com.ba/bhld/ (BH Telecom)
http://europronet.ba/bhld/ (Europronet – za korisnike cable zone: ftp://cableftp.europronet.ba/DOWNLOAD/OS/Linux/BHLD/)

Sva pitanja u vezi BHLDa i pomoć za korištenje možete dobiti na ULK forumu – sekcija BHLD koji redovno prati i sam dr Ribić (nick “megaribi”).

Pozivamo vas na promociju BHLD 2.0 koja će se održati u subotu 4. jula 2009. u 12:00 u Malom amfiteatru Elektrotehničkog fakulteta (ETF-MA):
* Mapa (koristite kontrolu sa lijeve strane da napravite zoom-in / zoom-out, ili vucite mišem mapu za skrolovanje)
Ovom prilikom ćete se moći učlaniti u Udruženje Linux korisnika, dobiti majice i kape sa logom Udruženja kao i odštampane BHLD priručnike sa CDom.

1. Udruženje Linux korisnika Bosne i Hercegovine
2. BHLD 2.0

More than six months from getting my CCNA certificate, I have passed CCNA Voice (few weeks ago actually, but I am not refreshing this blog as much as I would like to). For me, that was logical step to take, because I was working few years in VoIP industry and I was interested in Cisco’s way of solving some VoIP based tasks, like voice routing, productivity features (music on hold, call transfer, blind and consultative, after hours call blocking, directory, call forwarding, call park/pickup and so on). CCNA Voice cert is covering all of those topics in details including setup of Cisco Unity (their voicemail solution), codecs and many other configuration based things that you could face as real-world requirements (like PSTN fail-over for example). There is up to 65 questions on the test and you are having two hours for that. There is just one simulation on the exam (I have expected more, but there is no as much as on the CCNA exam). Questions are in the form of the single answer, multiple choice and drag and drop. It was real pleasure for me to prepare this exam, since I was having two deployments of Cisco VoIP in prior to my decision to get certified in this field, and at the moment I am dealing with Cisco voice gateways. In next few months I will try to get some more voice certs, depending on my free time. Everyone interested in voice over ip, or generally in voice and is Cisco oriented should check this huge and interesting area.

One of the good things about Spanning Tree Protocol is that it works well by default. I have faced multiple situations where i needed to deal with it, but it is rare that it does not preform like you would expect it to (also, I do remember one situation with Cisco WLS controller and Cisco switches from the series 500 / those are not Catalyst series and there is no console port / where I was trying to find solution while at the same time STP was blocking some ports and I was not able to find out what is really happening so i overnight at deployment).

Basically, STP is preventing evil infinite loops in our networks. And it works well. But there is occasion where it works against us, and that is situation where we have two interconnected switches with multiple physical connections. We would expect that if we have two separate connections between switches twice as much data could be sent from one switch to the other than if there was only one connection. But, in normal network scenario one of the ports based on certain criteria would be blocked in order to prevent infinite loops (broadcast storms).

So, to follow up the scenario, let say that we have two switches, switch1 and switch2 trunked on the FastEthernet0/4 and FastEthernet0/5, so if we execute:

switch1#show spanning vlan 20
Interface Role Sts Cost Prio.Nbr Type
—————- —- — ——— ——– ——————–
Fa0/4 Root FWD 19 128.11 P2p
Fa0/5 Altn BLK 19 128.12 P2p

we will see that FastEthernet0/4 is forwarding traffic (FWD state) and that FastEthernet0/5 i blocked (Role Alt and Sts BLK). So if we somehow reconfigure Fa0/5 to be in FWD state we could double the bandwidth available between the two switches if we could use that path that is currently being blocked.

Ok, to take advantage of this situation we could do two things:

1. To configure EtherChannel
2. To play with VLANS and STP costs and trunk allowed vlan command (if scenario allows that)

We will describe configuration of the EtherChannel. An Etherchannel is simply a logical bundling of 2 – 8 physical connections between two Cisco switches (it is interesting that EtherChannel is mentioned in Cisco CCNA curriculum as a technology but actually is not described from the configuration point).

To configure EtherChannel we will need to execute “channel-group 1 mode on” command on the trunked ports. We need to follow this procedure on all interconnected/trunked ports (because line protocol could stay in status down otherwise on ports).

Good thing is that STP will see EtherChannel as one virtual connection. If some connection in EtherChannel goes offline we will not face any STP recalculation, and of-course, transmission will be slowed but we will avoid STP recalculation and transmission delay .

Configuration as follows:

switch1#conf t
switch1(config)#interface fast 0/4
switch1(config-if)#channel-group 1 mode on
Creating a port-channel interface Port-channel 1

switch(config-if)#interface fast 0/5
switch1(config-if)#channel-group 1 mode on

switch2#conf t
switch2(config)#int fast 0/4
switch2(config-if)#channel-group 1 mode on
switch2(config-if)#int fast 0/5
switch2(config-if)#channel-group 1 mode on

To verify configuration we will use command:

switch2#show spanning vlan 20
Interface Role Sts Cost Prio.Nbr Type
—————- —- — ——— ——– —————
Po1 Desg FWD 12 128.65 P2p

Please note that instead of physical ports listed we are now able to see virtual port Po1 as the designated forwarding port. Po1 stands for Port-Channel1. It is the logical interface created automatically once the EtherChannel configuration is done.

Now, just for the comparison reasons, if one of the interconnected interfaces goes down in scenario without EtherChannel, STP will go through recalculation process and ports will go through learning, listening and other stated until it ends up in the FWD state. That process could take up to one minute! In case that we have EtherChannel in place, our Po1 listed above would still be in FWD state and we wouldn’t notice any transmission delay.

I strongly believe that we are dealing with really good technology and this is also one of the solutions (like HSRP) that could save a lot of time and make your day-to-day job easier.