Ing. William Castillo García: Bienvenidos

Ing. William Castillo García: Bienvenidos

CM process registration

Product : CATV
Part Involved : CM Motorola SB5101
Summary: Checking levels of upstream path
Scenario : Ensure return commands to HeadEnd
Company : ISP on Ecuador
Reference Date : September 2008.

The fordward (downstrean) is based on 256QAM and checking returns values through web tool monitoring CM levels, we can find next between modulation and SNR

QPSK 18dB
16QAM 21dB
64QAM 30dB
256QAM 33dB

Local ip address assigned for CM is 192.168.100.1 and ensure the RF cable is connected into CM before to turn it.

Remember that , CM has its HFC MAC (ID based on hexadecimal number) and MTA MAC is the ID assigned to MTA (HFC MAC + 2), here I add some terms related to this issue :

- After the begin ranging process (CM and CMTS confirm power level and what time should synchronize, calculate delay of the link, search the CMTS's slot and set up the schedule list). CM registered with CMTS.

- MTA send a "DHCP discovery"

- DHCP Send MTA an IP address, provisioning mode, MTA TFTP filename and other information in the DHCP options

- MTA asks DNS to resolve Provisioning Server FQDN

- DNS responds with IP address of Provisioning Server

- MTA requests its Configuration File from TFTP Server

- TFTP responds with Telephony Configuration File

- MTA sends Sys Log notification of provisioning complete (optional)

- MTA sends SNMP NOTIFY of provisioning complete to Provisioning Server (optional)

CMTS & CM

Product : CATV
Part Involved : CMTS BSR 64000
Summary: Connections involved from the HUB to the client
Scenario : Troubleshooting and following the best levels of RF
Company : ISP on Ecuador
Reference Date : August 2008.

Recomendations to follow :

- Always check the logic card configuration (domain) if this one was configured like 1x8 or 2x 1x4 and based on this will plug the test cable US / DS. For example ; talking about the configuration 2x 4x1 (with this ISP) DS0 is related to US 0,1,2,3 and DS1 is related to US 4,5,6,7, that is the downstream is loaded by 4 upstream.

- If CM was registered and this one is moved to several hubs, first connect the RF cable to CM and after that turn on it, otherwise the registration process takes more time because it is not going to find the downstream carrier will start to scan frecuencies.If CM does not get up, then check its status like as SNR levels, its probably the attenuation should be adjusted, and cables on bad states can cause communications problems between CM and CMTS

Take next as level of reference :
CMTS Tx = +55dBmV
CMTS Rx = +14dBmV (+/- 1dB)
CM Tx = 35 to 55 dBmV
CM Rx = 0 dBmV (+/- 5 dBmV)

Raid over VmWare

Product : Raid
Part Involved : Linux virtualization over VmWare
Summary: Management raid
Scenario : Testing & troubleshooting new features over linux
Company : National Institute of Cancer Sickness of Peru (INEN).
Reference Date : January 2008.

There are two videos that I have uploaded available taking this issue :
(1) Part 1 http://www.youtube.com/watch?v=gecy796j23o
(2) Part 2 http://www.youtube.com/watch?v=86n8rRyU29Y

and basically the command procedure follows the next steps :

> fdisk -l
> fdisk /dev/sdh
m (menu)
n (add new partition)
p (primary partition)
1
391 (cylinder 1-391)
p (print partition table)
L (list type of partitions available)
fd (Linux Raid Auto)
t (change system identification of partition)
L (List)
fd (Linux Raid Auto)
p (print partition table)
w (save & exist)
> fdisk /dev/sdi
> fdisk /dev/sdj
> ls /dev/md0 -l
> mdadm --create /dev/md1 --level=1 --raid-devices=2 /dev/sdi1 /dev/sdj1
> mkfs.ext3 /dev/md1
> mkdir /oracle/
> mount /dev/md1 /oracle/
> df -lh
> cp /var/log/* /oracle/ -rf
> ls /oracle/
> vi /etc/fstab
** /dev/md1 /oracle ext3 defaults 0 0
:wq!

Power Off Computer
Erase disk7, disk9, disk10
Power On Computer

> df -lh
> mdadm --monitor /dev/md1
> ls /oracle/
> mkdir /oracle/hola/
> fdisk -l
** J does not have partition
> fdisk /dev/sdj
m (menu)
p (primary partition)
1-261
t (change system identification of partition)
fd (Linux Raid Auto)
p (print partition table)
w (save & exit)
> mdadm --manage /dev/md1 --add /dev/sdj1
> fdisk -l : less

Dlink & uPnP

Product : dLink DCS 2100 & dLink DI-624
Part Involved : Port 1900 (UpNp)
Summary: With no appearently reason your LAN activity has increased a burst of broadcast
Scenario : There are some delays inserted into you LAN connections.
Company : National Cancer Institute of Peru (INEN).
Reference Date : February 2008 to April 2008.

There is a hidden knowledge that is only enough to assign an ip address an every equipment connect to our LAN connections, but what happen is this device saves to itself additional functions of routing tasks when at the begining was buyed to fullfill another function ?

But it works perfectly, after your plugged it on the outlet LAN (it is sure that you are going to forget it!), and that is according to my experience in the case of this two models, but monitored with the help of WinDump application in order to identify the device connected and the ports used, when and how.

I have attached two procedures videos to detect the activity to these ports related to this issues and the later steps to disable UpNp in the device.

I sure you, that this type broadcast will not become again, at least not for this reason !