[prev in list] [next in list] [prev in thread] [next in thread]
List: aix-l
Subject: Re: Antwort: Re: What is the way to observe network traffic on a RS/6000 Adapter?
From: Denis Garrido Mendes <Denis.Mendes () TECBAN ! COM ! BR>
Date: 2005-06-23 16:58:05
Message-ID: OF81335E2B.3D2C18CA-ON03257029.005C8C87-03257029.005D3415 () tecban ! com ! br
[Download RAW message or body]
Esta uma mensagem em v rias partes no formato MIME.
--=_alternative 005D340403257029_Content-Type: text/plain; charset="ISO-8859-1"
Content-Transfer-Encoding: quoted-printable
With de NMON.
http://www-128.ibm.com/developerworks/eserver/articles/nmon.html
Introduction
The nmon tool is designed for AIX and IBM eServer pSeries performance
specialists to use for analyzing AIX performance data, including the
following:
CPU utilisation
Memory use
Kernel statistics and run queue information
Disks I/O rates, transfers, and read/write ratios
Free space on file systems
Disk adapters
Network I/O rates, transfers, and read/write ratios
Paging space and paging rates
CPU and AIX specification
Top processors
IBM HTTP web cache
User defined disk groups
Machine details and resources
Async I/O
Workload Manager
ESS disks
NFS
Dynamic LPAR changes
Use this tool together with nmon analyser which loads the nmon output file
and automatically create dozens of graphs.
Denis Garrido Mendes
Suporte Técnico / CETEM
3244-8377
fmu@OERAG.DE
Enviado Por: IBM AIX Discussion List <aix-l@Princeton.EDU>
23/06/2005 13:42
Favor responder a
IBM AIX Discussion List <aix-l@Princeton.EDU>
Para
aix-l@Princeton.EDU
cc
Assunto
Antwort: Re: What is the way to observe network traffic on a RS/6000
Adapter?
Hm,
with tcpdump -i <interface>
I don't see the average network traffic and also no peak.
# tcpdump -i en0
tcpdump: listening on en0
18:34:29.711055834 charon_e.32851 > styx_e.hats.hermes_e: udp 76
18:34:29.712416916 styx_e.hats.hermes_e > medusa_e.hats.hermes_e: udp 76
18:34:29.763647374 phoenix_e.32862 > pegasus_e.hats.hermes_e: udp 76
18:34:29.928263936 ikarus_e.hats.hermes_e > charon_e.hats.hermes_e: udp 76
18:34:30.008060807 phoenix_e.hats.hermes_e > pegasus_e.hats.hermes_e: udp
76
Mark Ray
<raym@US.IBM.COM>
Gesendet von: IBM An
AIX Discussion aix-l@Princeton.EDU
List Kopie
<aix-l@Princeton.
EDU> Thema
Re: What is the way to observe
network traffic on a RS/6000
23.06.2005 18:05 Adapter?
Bitte antworten
an
IBM AIX
Discussion List
<aix-l@Princeton.
EDU>
Frank:
One of the better network traffic monitors is "tcpdump". The basic syntax
is 'tcpdump -i ,<interface>'
Note that, "netpmon" is more granular, but running it will incur a
significant performance hit on
your systems.
Mark Ray
IBM
fmu@OERAG.DE
Sent by: IBM AIX
Discussion List To
<aix-l@Princeton. aix-l@Princeton.EDU
EDU> cc
Subject
06/23/2005 11:53 What is the way to observe network
AM traffic on a RS/6000 Adapter?
Please respond to
IBM AIX
Discussion List
Hi,
on one of our SP node (AIX 5.2) runs a DB2 database (V8). All 120 clients
(Windows) works with this database and they are connect over 100MBit
Ethernet cards in our network.
Now we want the same application build again. But now the server is in
another location/city than the clients and we must work over a WAN and not
a LAN.
But I do not know the capacity which we need for the WAN or how big must
the line from city to city.
I reset the adapter statistik and check the output after 7h with entstat.
Here the result for en1 and en2:
-------------------------------------------------------------
ETHERNET STATISTICS (en1) :
Device Type: Gigabit Ethernet-SX PCI Adapter (14100401)
Hardware Address: 00:02:55:9a:54:e9
Elapsed Time: 0 days 7 hours 18 minutes 18 seconds
Transmit Statistics: Receive Statistics:
-------------------- -------------------
Packets: 1755437 Packets: 398349
Bytes: 2633851309 Bytes: 24161670
Interrupts: 5307 Interrupts: 358738
Transmit Errors: 0 Receive Errors: 0
Packets Dropped: 0 Packets Dropped: 0
Bad Packets: 0
Max Packets on S/W Transmit Queue: 77
S/W Transmit Queue Overflow: 0
Current S/W+H/W Transmit Queue Length: 0
-------------------------------------------------------------
ETHERNET STATISTICS (en3) :
Device Type: Gigabit Ethernet-SX PCI Adapter (14100401)
Hardware Address: 00:02:55:9a:54:e5
Elapsed Time: 0 days 2 hours 53 minutes 8 seconds
Transmit Statistics: Receive Statistics:
-------------------- -------------------
Packets: 1535981 Packets: 1067630
Bytes: 1853344547 Bytes: 199525044
Interrupts: 29289 Interrupts: 1042437
Transmit Errors: 0 Receive Errors: 0
Packets Dropped: 0 Packets Dropped: 0
Bad Packets: 0
Max Packets on S/W Transmit Queue: 105
S/W Transmit Queue Overflow: 0
Current S/W+H/W Transmit Queue Length: 3
When I calculte the bytes (for example Transmit) I get for en1 99 MByte/s
and for en3 56 MByte/s
I think that is very high and I want now control this values with a other
tool?
And this is the average and not a peak.
Is this possible?
Best regards,
Frank Mueller
--=_alternative 005D340403257029_Content-Type: text/html; charset="ISO-8859-1"
Content-Transfer-Encoding: quoted-printable
<br><font size=2 face="sans-serif">With de NMON.</font>
<br>
<br><font size=2 face="sans-serif">http://www-128.ibm.com/developerworks/eserver/articles/nmon.html</font>
<br>
<table width=100%>
<tr valign=top>
<td width=100%><font size=3>Introduction<br>
The </font><font size=3><tt>nmon</tt></font><font size=3> tool is designed
for AIX and IBM eServer pSeries performance specialists to use for analyzing
AIX performance data, including the following:</font>
<ul>
<li><font size=3>CPU utilisation </font>
<li><font size=3>Memory use </font>
<li><font size=3>Kernel statistics and run queue information </font>
<li><font size=3>Disks I/O rates, transfers, and read/write ratios </font>
<li><font size=3>Free space on file systems </font>
<li><font size=3>Disk adapters </font>
<li><font size=3><b>Network I/O rates, transfers, and read/write ratios
</b></font>
<li><font size=3>Paging space and paging rates </font>
<li><font size=3>CPU and AIX specification </font>
<li><font size=3>Top processors </font>
<li><font size=3>IBM HTTP web cache </font>
<li><font size=3>User defined disk groups </font>
<li><font size=3>Machine details and resources </font>
<li><font size=3>Async I/O </font>
<li><font size=3>Workload Manager </font>
<li><font size=3>ESS disks </font>
<li><font size=3>NFS </font>
<li><font size=3>Dynamic LPAR changes </font></ul></table>
<br>
<br>
<table width=100%>
<tr valign=top>
<td width=100%><font size=3><b>Use this tool together with </b></font><a \
href="http://www-106.ibm.com/developerworks/eserver/articles/nmon_analyser/index.html"><font \
size=3 color=blue><b><u>nmon analyser </u></b></font></a><font size=3><b>which loads \
the </b></font><font size=3><tt><b>nmon</b></tt></font><font size=3><b> output file \
and automatically create dozens of graphs.</b></font></table> <br>
<br><font size=2 face="sans-serif"><br>
<br>
Denis Garrido Mendes<br>
Suporte Técnico / CETEM<br>
3244-8377</font>
<br>
<br>
<br>
<table width=100%>
<tr valign=top>
<td width=40%><font size=1 face="sans-serif"><b>fmu@OERAG.DE</b> </font>
<br><font size=1 face="sans-serif">Enviado Por: IBM AIX Discussion List
<aix-l@Princeton.EDU></font>
<p><font size=1 face="sans-serif">23/06/2005 13:42</font>
<table border>
<tr valign=top>
<td bgcolor=white>
<div align=center><font size=1 face="sans-serif">Favor responder a<br>
IBM AIX Discussion List <aix-l@Princeton.EDU></font></div></table>
<br>
<td width=59%>
<table width=100%>
<tr>
<td>
<div align=right><font size=1 face="sans-serif">Para</font></div>
<td valign=top><font size=1 face="sans-serif">aix-l@Princeton.EDU</font>
<tr>
<td>
<div align=right><font size=1 face="sans-serif">cc</font></div>
<td valign=top>
<tr>
<td>
<div align=right><font size=1 face="sans-serif">Assunto</font></div>
<td valign=top><font size=1 face="sans-serif">Antwort: Re: What is the
way to observe network traffic on a RS/6000 Adapter?</font></table>
<br>
<table>
<tr valign=top>
<td>
<td></table>
<br></table>
<br>
<br>
<br><font size=2><tt>Hm,<br>
<br>
with tcpdump -i <interface><br>
I don't see the average network traffic and also no peak.<br>
<br>
<br>
# tcpdump -i en0<br>
tcpdump: listening on en0<br>
18:34:29.711055834 charon_e.32851 > styx_e.hats.hermes_e: udp 76<br>
18:34:29.712416916 styx_e.hats.hermes_e > medusa_e.hats.hermes_e: udp
76<br>
18:34:29.763647374 phoenix_e.32862 > pegasus_e.hats.hermes_e: udp 76<br>
18:34:29.928263936 ikarus_e.hats.hermes_e > charon_e.hats.hermes_e:
udp 76<br>
18:34:30.008060807 phoenix_e.hats.hermes_e > pegasus_e.hats.hermes_e:
udp<br>
76<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
Mark Ray
<br>
<raym@US.IBM.COM>
<br>
Gesendet von: IBM
An <br>
AIX Discussion
aix-l@Princeton.EDU
<br>
List
Kopie
<br>
<aix-l@Princeton.
<br>
EDU>
Thema <br>
Re:
What is the way to observe <br>
network
traffic on a RS/6000 <br>
23.06.2005 18:05
Adapter?
<br>
<br>
<br>
Bitte antworten
<br>
an
<br>
IBM AIX
<br>
Discussion List
<br>
<aix-l@Princeton.
<br>
EDU>
<br>
<br>
<br>
<br>
<br>
<br>
<br>
Frank:<br>
<br>
One of the better network traffic monitors is "tcpdump". The
basic syntax<br>
is 'tcpdump -i ,<interface>'<br>
Note that, "netpmon" is more granular, but running it will incur
a<br>
significant performance hit on<br>
your systems.<br>
<br>
Mark Ray<br>
IBM<br>
<br>
<br>
<br>
<br>
fmu@OERAG.DE<br>
Sent by: IBM AIX<br>
Discussion List
To<br>
<aix-l@Princeton.
aix-l@Princeton.EDU<br>
EDU>
cc<br>
<br>
Subject<br>
06/23/2005 11:53
What is the way to observe network<br>
AM
traffic on a RS/6000
Adapter?<br>
<br>
<br>
Please respond to<br>
IBM AIX<br>
Discussion List<br>
<br>
<br>
<br>
<br>
<br>
<br>
Hi,<br>
<br>
on one of our SP node (AIX 5.2) runs a DB2 database (V8). All 120 clients<br>
(Windows) works with this database and they are connect over 100MBit<br>
Ethernet cards in our network.<br>
<br>
Now we want the same application build again. But now the server is in<br>
another location/city than the clients and we must work over a WAN and
not<br>
a LAN.<br>
<br>
But I do not know the capacity which we need for the WAN or how big must<br>
the line from city to city.<br>
<br>
<br>
I reset the adapter statistik and check the output after 7h with entstat.<br>
Here the result for en1 and en2:<br>
<br>
-------------------------------------------------------------<br>
ETHERNET STATISTICS (en1) :<br>
Device Type: Gigabit Ethernet-SX PCI Adapter (14100401)<br>
Hardware Address: 00:02:55:9a:54:e9<br>
Elapsed Time: 0 days 7 hours 18 minutes 18 seconds<br>
<br>
Transmit Statistics:
Receive Statistics:<br>
--------------------
-------------------<br>
Packets: 1755437
Packets: 398349<br>
Bytes: 2633851309
Bytes: 24161670<br>
Interrupts: 5307
Interrupts: 358738<br>
Transmit Errors: 0
Receive Errors: 0<br>
Packets Dropped: 0
Packets Dropped: 0<br>
Bad Packets: 0<br>
Max Packets on S/W Transmit Queue: 77<br>
S/W Transmit Queue Overflow: 0<br>
Current S/W+H/W Transmit Queue Length: 0<br>
<br>
<br>
<br>
<br>
-------------------------------------------------------------<br>
ETHERNET STATISTICS (en3) :<br>
Device Type: Gigabit Ethernet-SX PCI Adapter (14100401)<br>
Hardware Address: 00:02:55:9a:54:e5<br>
Elapsed Time: 0 days 2 hours 53 minutes 8 seconds<br>
<br>
Transmit Statistics:
Receive Statistics:<br>
--------------------
-------------------<br>
Packets: 1535981
Packets: 1067630<br>
Bytes: 1853344547
Bytes: 199525044<br>
Interrupts: 29289
Interrupts: 1042437<br>
Transmit Errors: 0
Receive Errors: 0<br>
Packets Dropped: 0
Packets Dropped: 0<br>
Bad Packets: 0<br>
Max Packets on S/W Transmit Queue: 105<br>
S/W Transmit Queue Overflow: 0<br>
Current S/W+H/W Transmit Queue Length: 3<br>
<br>
<br>
When I calculte the bytes (for example Transmit) I get for en1 99 MByte/s<br>
and for en3 56 MByte/s<br>
I think that is very high and I want now control this values with a other<br>
tool?<br>
And this is the average and not a peak.<br>
<br>
Is this possible?<br>
<br>
<br>
<br>
Best regards,<br>
Frank Mueller<br>
</tt></font>
<br>
--=_alternative 005D340403257029_=--
[prev in list] [next in list] [prev in thread] [next in thread]
Configure |
About |
News |
Add a list |
Sponsored by KoreLogic