6509 CSM Connection Stats History

I have a 6509 with a CSM blade installed. I had a request to show conneciton statistics for a particular serverfarm for last week and how the connections were divied up between the servers. Are there any tools that can do this? I do have CiscoWorks but didn't see anything there. Any help would be greatly appreciated. Thanks.

You should be able to see the preconfigured reports. You will have to run the reports to generate the statistics.

Similar Messages

CSM connection stats missing?

Hi
I have a policy-map to insert an Ip address into the http header. I am also binding this to a sticky group using src ip address(See policy ICHAIN1-INSERT).
When I initiate a connection within the policy, although the connection is successful, when I try to view the connection stats (show mod csm 3 conns) my client is not listed in the connection table.
When I remove the "sticky-group 10" command then show the connection table, I see my client listed.
Is this normal behaviour?
Also when I issue the show sticky command the real ip's are garbled as shown:-
10 ip 143.52.208.19 D^OI^TD^Z^P!KP 65282
10 ip 10.6.1.14 P2^hD^Z^D^B*^X 79985
Config;
module ContentSwitchingModule 2
ft group 1 vlan 107
priority 20 alt 10
preempt
vlan 105 client
ip address 10.14.105.6 255.255.255.0
gateway 10.14.105.1
natpool CSM-PR1-USERS 10.14.105.10 10.14.105.18 netmask 255.255.255.0
probe ICHAIN-HTTP1 tcp
interval 10
failed 60
port 80
probe ZEN-APPS1-PRB tcp
interval 10
failed 60
port 524
probe ZEN-LDAP1-PRB tcp
interval 10
failed 60
port 636
probe ZEN-SERVER1-PRB tcp
interval 10
failed 60
port 524
map ICHAIN1-X-FOR header
insert protocol http header X-Forwarded-For header-value %is
real BONG
address 143.52.2.120
inservice
real HUORN
address 10.11.33.44
inservice
real ICHAIN101
address 10.14.72.21
inservice
real ICHAIN202
address 10.14.72.70
inservice
real JOSHUA
address 143.52.2.121
inservice
real KARAKA
address 143.52.2.42
inservice
real KARO
address 10.11.33.30
inservice
real PATE
address 10.11.33.32
inservice
serverfarm ICHAIN-BB1
nat server
nat client CSM-PR1-USERS
predictor leastconns
real name ICHAIN1
inservice
real name ICHAIN2
inservice
probe ICHAIN-HTTP1
serverfarm ZEN-APPS1
nat server
nat client CSM-PR1-USERS
predictor leastconns
real name BONG
inservice
real name JOSHUA
inservice
real name KARO
inservice
real name PATE
inservice
probe ZEN-APPS1-PRB
serverfarm ZEN-LDAP1
nat server
nat client CSM-PR1-USERS
predictor leastconns
real name HUORN
inservice
real name KARAKA
inservice
probe ZEN-LDAP1-PRB
serverfarm ZEN-SERVER1
nat server
nat client CSM-PR1-USERS
predictor leastconns
real name HUORN
inservice
real name KARAKA
inservice
probe ZEN-SERVER1-PRB
sticky 10 netmask 255.255.255.255 timeout 2880
policy ICHAIN1-INSERT
header-map ICHAIN1-X-FOR
sticky-group 10
serverfarm ICHAIN-BB1
vserver VIP-ICHAIN1
virtual 10.14.105.20 tcp www
serverfarm ICHAIN-BB1
sticky 2880 group 10
replicate csrp sticky
replicate csrp connection
persistent rebalance
slb-policy ICHAIN1-INSERT
inservice
vserver VIP-ICHAIN1-SSL
virtual 10.14.105.20 tcp https
serverfarm ICHAIN-BB1
sticky 1440 group 10
replicate csrp sticky
replicate csrp connection
persistent rebalance
inservice
vserver ZEN-APPS1-VIP
virtual 10.14.105.21 tcp 0
serverfarm ZEN-APPS1
replicate csrp connection
persistent rebalance
inservice
vserver ZEN-LDAP1-VIP
virtual 10.14.105.22 tcp 0
serverfarm ZEN-LDAP1
replicate csrp connection
persistent rebalance
inservice
vserver ZEN-SERVER1-VIP
virtual 10.14.105.23 tcp 0
serverfarm ZEN-SERVER1
replicate csrp connection
persistent rebalance
inservice
Many Thanks
Scott

Hello, I have the same problem running CSM Ver 4.1(7)
Showing connection stats with any of
show module csm 4 conn
show module csm 4 real
do not show any connections nor does it matter if I specify vserver or client or detail as the options
I am using cookie-based sticky and the show mod csm 4 sticky command seems to give correct info so that is not an issue for me

CSM Connection Stats

Hi all,
Just a quickie, we have implemented the CSM's into our portal and have 16 web servers in our webfarms. In a previous platform we used Alteon's which gave us a good indication for connections to each webserver, which we could then take out of service and ensure all sessions were off before working on the server. The CSM server connections seem very low (as we know thousands of connections come into our platform) and sometimes are in the single figures. Our platform is running HTTP in the main - so is there any way of finding out the HTTP active sessions - rather than HTTP connections
Cheers
James Robertson

James,
I'm not sure what you are looking for here.
An HTTP session would be 1 or more connections.
So, if your number of connections is low, the number of sessions would be even lower.
If there is no connection to a server, the server is not doing anything.
So the number of connections should be enough to decide which server is busy or not.
Gilles.

CSM Connections Timed-Out:

I have implemented a new application in our CSM. When I do a sho mod csm x stats I have noticed entrys in line Connections Timed-Out:
What does that mean?
If I do sho real sfarm I dont get any entrys for total conn failures
Regards
Mats

we have 2 timers per vserver.
idle timeout
pending timeout.
If a connection timed out it's because of one of these timers.
The idle timeout is if there is no traffic neither from client nor server.
The default is 1 hour.
The pending timeout is the max time allowed to complete the 3-way handshake.
The default is 10 sec.
These are not counted as failures.
A failure is when the server does not respond or respond with a reset.
Gilles.

Creating new Connections/Statements/ResultSets

Hi,
RDMS = Sybase
JDBC = Jconnect 5.5 obtained from Sybase website
Purpose:
To loop through the program below with different queries and produce
a resultset.
Details:
I have a program that creates a query stores it in a static setter.
The program calls the program below. The program below:
(1) obtains the query from a static getter
(2) creates a connection/statement/resultset
(3) returns to the calling program where
(a) resultset is displayed
(b) calls close() in the program below which closes the
connection/statement/resultset
Problem:
I thought that creating a new connection/statement/resultset
each time the program below is called would work.
The program below returns the correct resultset for the first query
but when called with a different query returns the first resultset that
was generated.
Question:
Do I have to use different objects for the connection/statement/resultset each time the program is called
even though the existing connection/statement/resultset is closed
before a new connection/statement/resultset is created ?
if yes,
I don't how to change the object for connection/statement/resultset
dynamically each time the program below is called.
Any assistance provided would be greatly appreciated !!
Thanks for your time,
YAM-SSM
package ecmutl;
     import java.awt.*;
     import java.sql.*;
     public class DbConnect implements Runnable {
          public static Thread queryThread = null;
          public static Font        fntF;
          public static Connection conn = null;
          public static ResultSet rs = null;
          public static Statement stmt = null;
          public static String nodename = EcmUtlLogicals.getNodeName();
          public static String password =     EcmUtlLogicals.getPassword();
          public String newline = "\n";
          public static String sqlstate;
          public static String message;
          public static String queryThreadDone = "false";
          public static String query = EcmUtlLogicals.getQuery();
          public static String dbname = EcmUtlLogicals.dbname;
          public static String username = EcmUtlLogicals.username;
             public static int SybaseReturnCode;
          void Connect() {
                    try {
                    Class.forName("com.sybase.jdbc2.jdbc.SybDriver").newInstance();
                       conn = DriverManager.getConnection("jdbc:sybase:Tds:"+nodename+"/"+dbname,username,password);
               catch (SQLException exc) {
                    sqlstate = exc.getSQLState();
                    message = exc.getMessage();
                    SybaseReturnCode = exc.getErrorCode();
               catch (InstantiationException iste) {
               catch (ClassNotFoundException cnfe) {
               catch (IllegalAccessException iae) {
               if (queryThread == null) {
                    queryThread = new Thread(this, "Query");
                    queryThread.start();
               } // close if
          } // close Connect method
          public void run() {
              try {
                                 stmt = DbConnect.conn.createStatement();
                    rs = stmt.executeQuery(query);
                    queryThread = null;
               } // close try
               catch (SQLException sqlex) {
                    sqlstate = sqlex.getSQLState();
                    message = sqlex.getMessage();
                    SybaseReturnCode = sqlex.getErrorCode();
                                queryThread = null;
                        } // close catch
                        catch (java.lang.Exception ex) {
                      ex.printStackTrace ();
               } // close catch
               queryThreadDone = "true";
          } // close run method
          static void close() {
               try {
                     if ( rs != null){
                          rs.close();
                     if ( stmt != null){
                          stmt.close();
                     if ( conn != null){
                          conn.close();
               } // close try
               catch (SQLException error){
          }// close close()
} // close DbConnect class

Then I'd say you should write a QueryThread object
that extends Thread and takes an SQL statement in its
constructor. In the run method, create a connection,
statement, and result set to run that query, close
them in reverse order when you're done, and provide a
getter to give access to the results. Don't return a
reference to the result set; put the results in a data
structure or CachedRowSet and close out the result set
as soon as you're done with it. ResultSets are
database cursors, which are scarce resources.
With this design, every thread has its own connection
and resources. They can't conflict with each other. -
MODHi MOD,
I finished coding a QueryThread object exactly as you instructed
and IT WORKS !!!!! The calling program calls QueryThread by:
QueryThread qt = new QueryThread(query);
qt.start();
QueryThread code:
package ecmutl;
     import java.awt.*;
     import java.sql.*;
     import java.util.List;
     import java.util.*;
     public class QueryThread extends Thread {
          public static Thread queryThread = null;
          public static Font        fntF;
          public static Connection conn = null;
          public static ResultSet rs = null;
          public static Statement stmt = null;
          public static String nodename = EcmUtlLogicals.getNodeName();
          public static String password = EcmUtlLogicals.getPassword();
          public String newline = "\n";
          public static String sqlstate;
          public static String message;
          public static String queryThreadDone = "false";
          public String query;
          public static String dbname = EcmUtlLogicals.dbname;
          public static String username = EcmUtlLogicals.username;
          public static int SybaseReturnCode;
          public QueryThread(String query) {
               this.query = query;
          public void run() {
               try {
                    Class.forName("com.sybase.jdbc2.jdbc.SybDriver").newInstance();
                    conn = DriverManager.getConnection("jdbc:sybase:Tds:"+nodename+"/"+dbname,username,password);
                    stmt = conn.createStatement();
                    rs = stmt.executeQuery(query);
                    List resultSetArray = new LinkedList();
                    resultSetArray = new ArrayList();
                    if (rs != null) {
                         int i;
                         ResultSetMetaData rsmd = rs.getMetaData ();
                         int numCols = rsmd.getColumnCount ();
                         for (i=1; i<=numCols; i++) {
                              if (i > 1) resultSetArray.add(",");
                              resultSetArray.add(rsmd.getColumnLabel(i));
                         resultSetArray.add(newline);
                         boolean more = rs.next ();
                         while (more) {
                              for (i=1; i<=numCols; i++) {
                                   if (i > 1) resultSetArray.add(",");
                                   resultSetArray.add(rs.getString(i));
                              resultSetArray.add(newline);
                              more = rs.next ();
                         EcmUtlLogicals.setResultSetArray(resultSetArray);
                         if (rs != null){
                              rs.close();
                         if (stmt != null) {
                              stmt.close();
                         if (conn != null) {
                              conn.close();
                    queryThread = null;
               catch (SQLException sqlex) {
                    sqlstate = sqlex.getSQLState();
                    message = sqlex.getMessage();
                    SybaseReturnCode = sqlex.getErrorCode();
                    queryThread = null;
               catch (java.lang.Exception ex) {
                                 ex.printStackTrace ();
               queryThreadDone = "true";
     }I have only be using Java for three months. Sometimes one needs a teacher to go along with the APIs.
Thanks for your patience/help/time
YAM-SSM

Purge Stats History taking too long

Hi,
I am doing a purge of stats history on my database and it is taking ages for what I believe should be a small data set
I have checked the stats history retention using
select dbms_stats.get_stats_history_availability from dual;
26/11/2010 22:10:42.668499000 +00:00
select sysdate-500
from dual;
12/12/2010 10:52:09
exec dbms_stats.purge_stats(sysdate-500);
The above purge has been running for over an hour ( and is still going ! ). This is on a test system with no real other activity going on within it.
Q1. Why is the above purge taking so long on what should be a relatively small data set ?
any help greatly appreciated,
Jim

So sorry,
from doc.
PURGE_STATS Procedure
This procedure purges old versions of statistics saved in the dictionary. To run this procedure, you must have the SYSDBA or both ANALYZE ANY DICTIONARY and ANALYZE ANY system privilege.
Syntax
DBMS_STATS.PURGE_STATS(
before_timestamp TIMESTAMP WITH TIME ZONE);
Delete all infor before 1,5 years. How many time you have beyond this date ??? the problem is the UNDO, purge using a batch process, delete from days to days, or using weeks.
HTH
Antonio NAVARRO

Hub 4 detailed connection stats??

Hello
Got my new hub 4 , all plugged in and working .However I can only get basic connection stats such as upload/download speed and time.
Where are detailed stats that I had with my hub 3??
Cheers
Oliver

I read the parts of the linked chat thread regarding disabling the noise/attenuation stats, and the reasoning for leaving it out does not make sense at all. Users are not going to lose sleep over extra technical information in the "Advanced" section - they know it's there for techincal support if they need it.
Heck, you could even completely hide it by not having a clickable link anywhere and only tell people about the HomeHub "special" link on the forums or BT support - then people won't get confused if they click where they shouldn't.
There are plenty of posts on these forums from people with older HomeHubs where they copy their stat page and people on the forum are able to easily identify what is going wrong (noise margin, etc). Easy for everyone! It is now made much more difficult to dig through logs to find this information now, even for a competent user like myself (I'm a software engineer) who isn't a fan of digging through logs if I can help it.
I think this really needs to be reconsidered in future firmware versions because - in my opinion - it is insane to completely disable something like this. It certainly wasted a lot of my time looking for it when I expected it to be there.

SIP 6945 Keypad stops working after Connected state

I have a lot of 6945s/3905s/6961s all working fine.
However I have 1 6945 phone that after dialing a number the Keypad stops working. So during the Connected state the keypad stop working. Ring Out is fine.
Any ideas? Bad hardware?
6945 SIP
9-3-3-2
CallManager
9.1.1.21018-1

When reading my report, I forgot to mention how the external hard drive is formatted :
Hard Drive FW/USB:
Manufacturer: LaCie
Model: 0x0
GUID: 0xD04B78070AD64B
Maximum Speed: Up to 400 Mb/sec
Connection Speed: Up to 400 Mb/sec
Sub-units:
Hard Drive FW/USB Unit:
Unit Software Version: 0x10483
Unit Spec ID: 0x609E
Firmware Revision: 0x110
Product Revision Level:
Sub-units:
Hard Drive FW/USB SBP-LUN:
Capacity: 298.09 GB
Removable Media: Yes
BSD Name: disk1
Mac OS 9 Drivers: No
Partition Map Type: APM (Apple Partition Map)
S.M.A.R.T. status: Not Supported
Volumes:
Aristotle:
Capacity: 297.97 GB
Available: 164.85 GB
Writable: Yes
File System: Journaled HFS+
BSD Name: disk1s3
Mount Point: /Volumes/Aristotle
Raf

-24994 Runtime environment error. 1,wrong connection state.

Dear All,
While i was restoring the log files on a server from DBM Gui 7.6.0.
I got this error
-24994 Runtime environment error [backup_restore_replace "log" "B:\Backup from PRD\PRD_LOG "2974]; 1, wrong connection state.
The size of the next four files after the PRD_LOG.2974 is '0' since the backups scheduled were failed for lack of space.
What should i do in this situation?
Thanks.
Regards,
Mudassir.Imtiaz

Now when i tried to start it all over again.
Its giving me the following error.
-24988 Sql error [backup_restore "log" LOG 2975]; -903, Host file I/O error.
Any suggestions how to get over this.
Thanks.
Regards,
Mudassir.Imtiaz

Connect statement with input string in a script

I would like to enter a password for each connect statement after I start my.sql
my.sql
connect scott/&p_password@orcl;
grant SELECT to peter on table_1;
SQL>@my.sql
SQL> connect scott/&p_password@orcl;
Please enter a value for p_password: grant SELECT to peter on table_1;
invalid option.
using: CONNECT <username> [AS SYSDBA|SYSOPER]
SQL> ...
Sql accept the grant statement after the connect statement as an input value.
What must I change in my script???
Best regards,
Tom Rakete
Message was edited by:
Tom Rakete

Are you trying to start your script from within SQL*Plus?
Afraid no. Because the prompt "Please enter a value for ..." sounds suspicious.
C:\my.sql
conn max/&p_password@pdoxora
grant select on test to public;Copy-paste from SQL*Plus:
SQL> @C:\my.sql
Enter value for p_password: *******
Connected.
Grant succeeded.In other words, SQL*Plus prompts "Enter value", not "Please enter a value".
So what tool are you using?
Regards,
Andrew.

SPA525G2 - Can't see pressed keys/ chars - Connected State

I have CISCO SPA525G2 already deployed at several clients premises (FW 7.5.1a).
Unfortunately we’ve found a blocking issue that we hope to solve faster:
Fail to show dialed alphanumeric characters in the Connected State, eg:
IVR interaction don’t displays the characters pressed and heard by the user
Conference call PIN isn’t showed when the characters are pressed and heard by the user
I’ve tried to change configurations on customized soft keys but no results…current configuration is:
<Connected_Key_List ua="na">hold|1;endcall|2;xfer|3;bxfer|4;conf;crdstart;crdstop;xferLx;confLx;park;phold;flash;</Connected_Key_List>
Last testing was with 7.5.4 firmware and no results.
If this is firmware related the only thing that you need it’s to copy the behavior of on HOLD to show the line with chars as you press some key.
Does anyone from CISCO could help with this ? I've already opened a case at https://supportforums.cisco.com/community/netpro/small-business/onlinechatsupport and no response...
Thanks in advance,
João Silva

This worked for me - it turned out I had McAfee and Windows firewall enabled - I disabled windows firewall and added all IPs in the safe zone on my McAfee firewall. For example: 192.168.1.1 to 192.168.1.254. You could also try to just disable all firewalls for now.
Next step - I kept only one computer plugged into the Linksys at a time (physically unplugged them from the router inputs/shutdown wireless). Then do the network setup on each machine:
Programs-Accesories-Communication-Network Setup Wizard
Click next to all screens until you get to:
When it asks you "Do you want to use shared connection" pick "No, let me choose another way to connect to internet", pick "This computer connect to the internet through a residential gateway", give the computer a name, hit next and call the network something, say HOME. FInish and repeat the same for all machines (you can create a network CD/floppy with the settings). Make sure only 1 machine connected to the Linksys at a time. Keep the name HOME on all other setups.
This fixed it for me.

Cisco ACE - CLSRST connection state

hi,
During performance testing with IBM Rational Performance Tester tool with 500 users, I noticed above 4000 connections in 'show conn' display and most of the connection states showed as 'CLSRST'.
On doing search the meaning I found is
'CLSRST-ACE closed the connection by resetting it'
What does it mean and is it normal to see so many CLSRST connections.
Thanks.

It looks as if you are hitting CSCsw83500.
Syed Iftekhar Ahmed

Where did we find connection statement interfaces?

where did we find connection ,statement interfaces when we are working with type 4 driver?

where did we find connection ,statement interfaces
when we are working with type 4 driver?The interfaces are always present in rt.jar. :)

Purging Stats History Tables

Hi,
I am trying to purge down the following Stats History tables per note 1055547.1
SYS.WRI$_OPTSTAT_HISTHEAD_HISTORY
SYS.WRI$_OPTSTAT_HISTGRM_HISTORY
This note and many others suggest using
exec dbms_stats.purge_stats(sysdate-<days>);
and making the <days> such that you are getting bite size chunks for the purge so not to create problems with my Undo tablespace.
My problem is that I have 500 days worth of stats history to get rid of !
Any suggestions how to achieve this in a more efficient manner than running the above purge_stats command lots of times ?
thanks,
Jim

Any suggestions how to achieve this in a more efficient manner than running the above purge_stats command lots of times ? wrap it in a day-by-day loop.
do not forget to clean up orphaned STATS$SQLTEXT rows in every cycle;

I do not understand a v$sqlarea statement history

Dears,
I collect a history of SQL statements from Oracle 8i server. I use a statement like
SELECT first_load_time,sql_text FROM v$sqlarea
I get these rows:
=======================================================
11:23:15 SELECT DocumentID,DocTitul, Documents.ROWID FROM Documents WHERE DocumentID IN (SELECT distinct(DocumentID) FROM StoreHistory WHERE OperationType IN (6,7) AND Status = 'A') AND Status = 'A' AND SafeLevel <= 5 ORDER BY DocTitul
=======================================================
11:23:15 SELECT DocumentID,DocTitul, Documents.ROWID FROM Documents WHERE DocumentID IN (SELECT distinct(DocumentID) FROM StoreHistory WHERE OperationType IN (:V0001,:V0002) AND Status = 'A') AND Status = 'A' AND SafeLevel <= 5 ORDER BY DocTitul
=======================================================
11:23:15 SELECT DocumentID,DocTitul, Documents.ROWID FROM Documents WHERE DocumentID IN (SELECT distinct(DocumentID) FROM StoreHistory WHERE OperationType IN (6,:V0001) AND Status = 'A') AND Status = 'A' AND SafeLevel <= 5 ORDER BY DocTitul
=======================================================
I have two questions. Why there are ":V0001" and ":V0002" variables? Can I switch off such behaviour? I need, that there will be exact queries I send to server in the v$sqlarea dynamic view.
Also, I get these rows:
=======================================================
11:22:47 SELECT COUNT(*) FROM SysRights WHERE ActionPath = 'STORE/EVID' AND Enabled = 1 AND (UserID = 19)
=======================================================
11:22:47 SELECT COUNT(*), SysRights.ROWID FROM SysRights WHERE ActionPath = 'STORE/EVID' AND Enabled = 1 AND (UserID = 19)
=======================================================
I have another question. Why there is a "SysRights.ROWID" column on the query? Original query with agregate function does not have such column, of course.
Thank you for any information,
Richard
PS: Server (8i) is running on WinNT 4.0. Client (8.0.5[??? I am not sure]) is running on Win2000. Connect string contains "Provider=MSDAORA;".

Richard
:V0001 and :V0002 are bind variables or place holders. They indicate that the query comes from an application (eg Forms) that is using variables to hold values. This behaviour is very useful because it allows the database to reuse parsed queries and stored execution plans for queries with an identical structure. This is very efficient. If you were to switch off the use of bind variables then the database will have to parse and plan each query you send it, which will havw a deleterious affect on performance.
Some of this behaviour you can't override (eg the Forms stuff) but with SQL and PL/SQL you can (in 8.1.6+) by setting this parameter CURSOR_SHARING=EXACT
HTH, APC

6509 CSM Connection Stats History

Similar Messages

Maybe you are looking for