Re: [LARTC] Graphing tc output

2003-12-11 Thread Michal Charvat
here is output from snmpwalk on my server:
snmpwalk -m GNU-LINUX-KERNEL-QOS -c public localhost qos

GNU-LINUX-KERNEL-QOS::qosDeviceIndex.0.0 = INTEGER: 0
GNU-LINUX-KERNEL-QOS::qosDeviceIndex.1.65536 = INTEGER: 1
GNU-LINUX-KERNEL-QOS::qosDeviceIndex.5.65536 = INTEGER: 5
GNU-LINUX-KERNEL-QOS::qosDeviceIndex.5.65547 = INTEGER: 5
GNU-LINUX-KERNEL-QOS::qosDeviceIndex.5.65548 = INTEGER: 5
GNU-LINUX-KERNEL-QOS::qosDeviceIndex.5.65549 = INTEGER: 5
GNU-LINUX-KERNEL-QOS::qosDeviceIndex.5.65550 = INTEGER: 5
GNU-LINUX-KERNEL-QOS::qosDeviceIndex.5.131072 = INTEGER: 5
GNU-LINUX-KERNEL-QOS::qosDeviceIndex.5.196608 = INTEGER: 5
GNU-LINUX-KERNEL-QOS::qosDeviceIndex.5.262144 = INTEGER: 5
GNU-LINUX-KERNEL-QOS::qosDeviceIndex.5.327680 = INTEGER: 5
GNU-LINUX-KERNEL-QOS::qosHandle.0.0 = Gauge32: 0
GNU-LINUX-KERNEL-QOS::qosHandle.1.65536 = Gauge32: 65536
GNU-LINUX-KERNEL-QOS::qosHandle.5.65536 = Gauge32: 65536
GNU-LINUX-KERNEL-QOS::qosHandle.5.65547 = Gauge32: 65547
GNU-LINUX-KERNEL-QOS::qosHandle.5.65548 = Gauge32: 65548
GNU-LINUX-KERNEL-QOS::qosHandle.5.65549 = Gauge32: 65549
GNU-LINUX-KERNEL-QOS::qosHandle.5.65550 = Gauge32: 65550
GNU-LINUX-KERNEL-QOS::qosHandle.5.131072 = Gauge32: 131072
GNU-LINUX-KERNEL-QOS::qosHandle.5.196608 = Gauge32: 196608
GNU-LINUX-KERNEL-QOS::qosHandle.5.262144 = Gauge32: 262144
GNU-LINUX-KERNEL-QOS::qosHandle.5.327680 = Gauge32: 327680
GNU-LINUX-KERNEL-QOS::qosDeviceName.0.0 = 
GNU-LINUX-KERNEL-QOS::qosDeviceName.1.65536 = STRING: lo
GNU-LINUX-KERNEL-QOS::qosDeviceName.5.65536 = STRING: eth0
GNU-LINUX-KERNEL-QOS::qosDeviceName.5.65547 = STRING: eth0
GNU-LINUX-KERNEL-QOS::qosDeviceName.5.65548 = STRING: eth0
GNU-LINUX-KERNEL-QOS::qosDeviceName.5.65549 = STRING: eth0
GNU-LINUX-KERNEL-QOS::qosDeviceName.5.65550 = STRING: eth0
GNU-LINUX-KERNEL-QOS::qosDeviceName.5.131072 = STRING: eth0
GNU-LINUX-KERNEL-QOS::qosDeviceName.5.196608 = STRING: eth0
GNU-LINUX-KERNEL-QOS::qosDeviceName.5.262144 = STRING: eth0
GNU-LINUX-KERNEL-QOS::qosDeviceName.5.327680 = STRING: eth0
GNU-LINUX-KERNEL-QOS::qosType.0.0 = 
GNU-LINUX-KERNEL-QOS::qosType.1.65536 = STRING: sfq
GNU-LINUX-KERNEL-QOS::qosType.5.65536 = STRING: htb
GNU-LINUX-KERNEL-QOS::qosType.5.65547 = STRING: htb
GNU-LINUX-KERNEL-QOS::qosType.5.65548 = STRING: htb
GNU-LINUX-KERNEL-QOS::qosType.5.65549 = STRING: htb
GNU-LINUX-KERNEL-QOS::qosType.5.65550 = STRING: htb
GNU-LINUX-KERNEL-QOS::qosType.5.131072 = STRING: sfq
GNU-LINUX-KERNEL-QOS::qosType.5.196608 = STRING: sfq
GNU-LINUX-KERNEL-QOS::qosType.5.262144 = STRING: sfq
GNU-LINUX-KERNEL-QOS::qosType.5.327680 = STRING: sfq
GNU-LINUX-KERNEL-QOS::qosBytes.0.0 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosBytes.1.65536 = Counter32: 133920
GNU-LINUX-KERNEL-QOS::qosBytes.5.65536 = Counter32: 69312599
GNU-LINUX-KERNEL-QOS::qosBytes.5.65547 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosBytes.5.65548 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosBytes.5.65549 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosBytes.5.65550 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosBytes.5.131072 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosBytes.5.196608 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosBytes.5.262144 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosBytes.5.327680 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosPackets.0.0 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosPackets.1.65536 = Counter32: 1461
GNU-LINUX-KERNEL-QOS::qosPackets.5.65536 = Counter32: 547924
GNU-LINUX-KERNEL-QOS::qosPackets.5.65547 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosPackets.5.65548 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosPackets.5.65549 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosPackets.5.65550 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosPackets.5.131072 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosPackets.5.196608 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosPackets.5.262144 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosPackets.5.327680 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosDropped.0.0 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosDropped.1.65536 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosDropped.5.65536 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosDropped.5.65547 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosDropped.5.65548 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosDropped.5.65549 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosDropped.5.65550 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosDropped.5.131072 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosDropped.5.196608 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosDropped.5.262144 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosDropped.5.327680 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosOverlimit.0.0 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosOverlimit.1.65536 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosOverlimit.5.65536 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosOverlimit.5.65547 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosOverlimit.5.65548 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosOverlimit.5.65549 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosOverlimit.5.65550 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosOverlimit.5.131072 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosOverlimit.5.196608 = Counter32: 0
GNU-LINUX-KERNEL-QOS::qosOverlimit.5.262144 = Counter32: 0

Re: [LARTC] Graphing tc output

2003-12-11 Thread Stef Coene
On Thursday 11 December 2003 13:57, Michal Charvat wrote:
 Hi,

 Sorry for that, but yet another update :-)

 Originally I compiled code on RH8.0, now I tried on Debian stable,
 and I had to make some changes (looks like older gcc didn't like
 my code :-( ).
It's working :)
It's time to generate some graphs .

Stef

-- 
[EMAIL PROTECTED]
 Using Linux as bandwidth manager
 http://www.docum.org/
 #lartc @ irc.openprojects.net

___
LARTC mailing list / [EMAIL PROTECTED]
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/


Re: [LARTC] Graphing tc output

2003-12-11 Thread Stef Coene
On Thursday 11 December 2003 14:23, Stef Coene wrote:
 On Thursday 11 December 2003 13:57, Michal Charvat wrote:
  Hi,
 
  Sorry for that, but yet another update :-)
 
  Originally I compiled code on RH8.0, now I tried on Debian stable,
  and I had to make some changes (looks like older gcc didn't like
  my code :-( ).

 It's working :)
 It's time to generate some graphs .
Mhh.  It should be handy if there is a way to determine the parent-child 
relation ship.

Stef

-- 
[EMAIL PROTECTED]
 Using Linux as bandwidth manager
 http://www.docum.org/
 #lartc @ irc.openprojects.net

___
LARTC mailing list / [EMAIL PROTECTED]
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/


Re: [LARTC] Graphing tc output

2003-12-11 Thread Michal Charvat
 Mhh.  It should be handy if there is a way to determine the parent-child
 relation ship.

You are right, handles in this case are a bit confusing (major:minor
together isn't the best way...). You can determine parent-child
relationship from these numbers (as said in archive, higher 16bits are
major number, lower minor).

It would be possible to provide major/minor separately in table and leave
indexing in major:minor scheme as is now.
What do you think about this possibility?

Michal
___
LARTC mailing list / [EMAIL PROTECTED]
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/


Re: [LARTC] Graphing tc output

2003-12-11 Thread Stef Coene
On Thursday 11 December 2003 16:19, Michal Charvat wrote:
  Mhh.  It should be handy if there is a way to determine the parent-child
  relation ship.

 You are right, handles in this case are a bit confusing (major:minor
 together isn't the best way...). You can determine parent-child
 relationship from these numbers (as said in archive, higher 16bits are
 major number, lower minor).

 It would be possible to provide major/minor separately in table and leave
 indexing in major:minor scheme as is now.
 What do you think about this possibility?
That's not what I wanted to say.  I wanted to know if you can entry a new 
entry that gives all the classes (or qdisc if the parent is a leaf class) 
attached to an other class.  Like
SNMPv2-SMI::enterprises.18756.1.1.12.3.65536 = STRING: 10:10 10:11

Stef

-- 
[EMAIL PROTECTED]
 Using Linux as bandwidth manager
 http://www.docum.org/
 #lartc @ irc.openprojects.net

___
LARTC mailing list / [EMAIL PROTECTED]
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/


Re: [LARTC] Graphing tc output

2003-12-11 Thread Michal Charvat
I can add a new entry, that won't be a problem.
Parent string should be easy to add, but children don't know right
now how to do that. I'll try...

Michal Charvat
Prokon R.B.S. Louny v.o.s.
___
LARTC mailing list / [EMAIL PROTECTED]
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/


Re: [LARTC] Graphing tc output

2003-12-11 Thread Stef Coene
On Thursday 11 December 2003 17:27, Michal Charvat wrote:
 I can add a new entry, that won't be a problem.
 Parent string should be easy to add, but children don't know right
 now how to do that. I'll try...
Parent or child string.  If I (we) know one, we can find the other.

Stef

-- 
[EMAIL PROTECTED]
 Using Linux as bandwidth manager
 http://www.docum.org/
 #lartc @ irc.openprojects.net

___
LARTC mailing list / [EMAIL PROTECTED]
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/


RE: [LARTC] Graphing tc output

2003-12-10 Thread Chijioke Kalu
Hi

I use Stef Coene graphing tool see docum.org very good also check out 
bandwidthd @ sf.net, another nice tool that shows you bandwidth usage per 
ip.

And to top it off, for your whole interface, plug in mrtg with rrd nice look

Kalu

Hi,

My HTB shaping is working perfectly, and i have setup a bridge and hosts 
behind the bridge are now getting only the bandwidth they are supposed to 
get. Thanks to every one here for helping me out :)

Now I am thinking of making a graph of traffic going through various 
classes. Is their any thing readily available for it, or is parsing the 
output if tc and passing it to rrdtool the preferred way? Any suggestions?

Thanks,

raj
___
LARTC mailing list / [EMAIL PROTECTED]
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/
_
Take advantage of our best MSN Dial-up offer of the year — six months 
@$9.95/month. Sign up now! http://join.msn.com/?page=dept/dialup

___
LARTC mailing list / [EMAIL PROTECTED]
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/


Re: [LARTC] Graphing tc output

2003-12-10 Thread Michal Charvat
Hi,
I've recently written snmp extension to Net-SNMP package
(www.net-snmp.net), where it's possible to get kernel statistics directly
via snmp requests.

But it's still beta quality :-( (known one memory leak, I'm going to look
at it tommorow)

If you are interested, I can put it somewhere...

Michal Charvat
Prokon R.B.S. Louny v.o.s.


On Wed, 10 Dec 2003, Rajkumar S wrote:

 Hi,

 My HTB shaping is working perfectly, and i have setup a bridge and hosts
 behind the bridge are now getting only the bandwidth they are supposed
 to get. Thanks to every one here for helping me out :)

 Now I am thinking of making a graph of traffic going through various
 classes. Is their any thing readily available for it, or is parsing the
 output if tc and passing it to rrdtool the preferred way? Any suggestions?

 Thanks,

 raj
 ___
 LARTC mailing list / [EMAIL PROTECTED]
 http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/

___
LARTC mailing list / [EMAIL PROTECTED]
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/


Re: [LARTC] Graphing tc output

2003-12-10 Thread Damjan
 I've recently written snmp extension to Net-SNMP package
 (www.net-snmp.net), where it's possible to get kernel statistics directly
 via snmp requests.
 
 But it's still beta quality :-( (known one memory leak, I'm going to look
 at it tommorow)
 
 If you are interested, I can put it somewhere...

Sure everybody is interested :)

put it somewhere, maybe someone will find that memory leak before you
do :)


-- 
Damjan Georgievski
jabberID: [EMAIL PROTECTED]
___
LARTC mailing list / [EMAIL PROTECTED]
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/


Re: [LARTC] Graphing tc output

2003-12-10 Thread Michal Charvat
Ok, I've put my code to: http://x-ray.prokon.cz/data/snmp.tgz

It's quite simple, some code is copied from tc utility source... follow
instructions in package and it should work...

Michal Charvat
Prokon R.B.S. Louny v.o.s.


On Wed, 10 Dec 2003, Damjan wrote:

  I've recently written snmp extension to Net-SNMP package
  (www.net-snmp.net), where it's possible to get kernel statistics directly
  via snmp requests.
 
  But it's still beta quality :-( (known one memory leak, I'm going to look
  at it tommorow)
 
  If you are interested, I can put it somewhere...

 Sure everybody is interested :)

 put it somewhere, maybe someone will find that memory leak before you
 do :)


 --
 Damjan Georgievski
 jabberID: [EMAIL PROTECTED]
 ___
 LARTC mailing list / [EMAIL PROTECTED]
 http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/

___
LARTC mailing list / [EMAIL PROTECTED]
http://mailman.ds9a.nl/mailman/listinfo/lartc HOWTO: http://lartc.org/