Multiple instances of EIGRP or static routes

I'm building a network which needs to have All but one of it's private networks pass through a DMVPN, all the routes are advertised through EIGRP, that part works great!
I have a private VLAN that only has access onto the internet, the address is Nat'ed over to a public IP address. Each router, there's six of them, are neighbors to two other routers. The furthest router to the internet has to go through three routers to get to the internet. My current idea is to use static routes on all the routers to the Internet gateway router. Then let recursive routing sort out each hop. What I would rather do is have EIGRP do all that. I really don't want to mess with the EIGRP that's running for the DMVPN tunnels, I'd like to have another instance of EIGRP run on the routers that will route the users to the Internet.
Does anyone have any thoughts concerning this design.
Thanks.
Mitch

Mitch
I am not clear about what you are attempting to achieve and not very clear about the topology. So my answer may or may not be on target. If it is not perhaps you can help us understand a little better what is involved.
I believe that what you are saying is that you have an existing network with multiple locations connected over DMVPN and that you run EIGRP as the routing protocol for that network. I believe you are also saying that there is one network segment which needs access to the Internet but should not be able to access the other parts of your network.
You say that the address of this other segment is NATed but are not clear whether the translation is ont the router where the segment is located or is on the Internet gateway router.
Probably the traditional solution for this would be to provide a default route for this segment pointing toward the Internet gateway router, to have a route on the Internet gateway router (and other routers along the path toward where the network is located), and a series of access lists on each router along the way which allows passage to the Internet and denies access to local resources.
I would propose a somewhat different solution. I believe that it would work if you configure a GRE tunnel between the router where the segment is located and the Internet Gateway router. On the router where the segment is located you could do Policy Based Routing to send traffic from the private segment to the Internet over the GRE tunnel (which effectively isolates it from your other resources). You might want Policy Based Routing on the Internet gateway router to be sure that traffic from the private segment was forwarded only to the Internet (though you might not need that). The Internet gateway router could have a route (probably a static route) which sends traffic to the private segment over the GRE tunnel.
Let us know what you think of this. And if it is off the mark perhaps you could clarify a bit.
HTH
Rick

Similar Messages

Static member problem in multiple instance of web application

Problem
I am working on a product which is java based. I have a war file. I have to deploy this war file in JBoss. Further, I need to run multiple instance of this war file in same JBoss instance.
Issue
     In that war file there is servlet Start.java. That servlet calls an other class from its init method. Here is code
public class Start extends HttpServlet
     public void init(ServletConfig config) throws ServletException
          super.init(config);
          DB.loadProperties (�db.properties�);
     } // init
     public void doGet(HttpServletRequest request, HttpServletResponse response)
          throws ServletException, IOException
               } // doGet
     public void doPost(HttpServletRequest request, HttpServletResponse response)
          throws ServletException, IOException
               } // doPost
} // Start
public final class DB implements Serializable
     public static final String     P_UID = "ApplicationUserID";
     private static final String     DEFAULT_UID = "System";
     public static final String     P_PWD = "ApplicationPassword";
     private static final String     DEFAULT_PWD ="System";
     public static final String     P_STORE_PWD = "StorePassword";
     private static Properties           s_prop = new Properties();
public static boolean loadProperties (String filename)
     {    // looks for properties file
s_prop = new Properties();
          FileInputStream fis = null;
          try
               fis = new FileInputStream(filename);
               s_prop.load(fis);
               fis.close();
          catch (FileNotFoundException e)
               log.warning(filename + " not found");
               loadOK = false;
     �}
As you can see if I deployed two instance of this war in JBoss and assume each instance point to its own properties file. The �Static� members are problem.
Restriction
     I am not allowed change the code or logic of the product. I must be able to get some thing outside or apply a patch to solve this problem.
Please Help
Possible option that I have thought
1 Using multiple JVM - Not an option in this project
2 Using multiple JBoss � Not the requirement (need to run only on single Instance of JBoss)
3 Use own class loader for war/ear � don�t know how in JBoss need help.
I am using JBoss 3.2.3
Thanks,
Kumaran

Static members belong to the class. So, you needto
think of multiple JVMs for your problem.Thanks for reply.
don�t know how in JBoss need helpJust guessing... clustered environment MAYBE
RichThanks Rich,
I don�t know how Cluster environment solve my problem. I need more information on your thought. Cloud you explain it.
Kumaran

How do you Redistribution EIGRP into OSPF and maintain a distance of 250 for a static route?

Ok, I have scoured the forums long enough and have to post. The design is below. I moved a firewall to our new data center, which required adding some static routes for VPN connections and broadband backups. To minimize the amount of static routes I redistribute static into EIGRP with a route-map and prefix-list.
My problem is the next part of my network. When the data leaves my 56128's it hits an edge device connecting to our dark fiber. On this edge device I am running OSPF onto the dark fiber, then redistribute some EIGRP subnets into OSPF and again all is well.
Everything works up until the point the redistributed routes hit my RIB at my main data center where I am running IBGP. IBPG is run between our MPLS router and core for all our remote sites. When my backup route from the 56128's hits the cores, it supersedes the BGP route because the AD route O E2 [110/20] is lower than the BGP AD B [200/0]. Given the configuration below what can be done to remedy this? Oh when I redistribute I can only change the AD for the backup routes, all other routes should stay the same.
56128's where my static routes are:
ip route 192.168.101.0/24 192.168.30.77 name firewall 250
router eigrp 65100
redistribute static route-map Static-To-Eigrp
route-map Static-To-Eigrp permit 10
match ip address prefix-list Static2Eigrp
ip prefix-list Static2Eigrp seq 2 permit 192.168.101.0/24
Edge device:
router eigrp 65100
network 172.18.0.5 0.0.0.0
network 172.18.0.32 0.0.0.3
network 172.18.0.36 0.0.0.3
redistribute ospf 65100 metric 2000000 0 255 1 1500
redistribute static metric 200000 0 255 1 1500 route-map STATICS_INTO_EIGRP
passive-interface default
no passive-interface Port-channel11
no passive-interface Port-channel12
eigrp router-id 172.18.0.5
router ospf 65100
router-id 172.18.0.5
log-adjacency-changes
redistribute eigrp 65100 subnets route-map EIGRP_INTO_OSPF
passive-interface default
no passive-interface GigabitEthernet1/0/1
no passive-interface GigabitEthernet1/0/2
no passive-interface GigabitEthernet2/0/1
no passive-interface GigabitEthernet2/0/2
network 172.18.0.0 0.0.255.255 area 0
ip prefix-list EIGRP_INTO_OSPF seq 5 permit 172.18.0.0/16 le 32
ip prefix-list EIGRP_INTO_OSPF seq 10 permit 192.168.94.0/29 le 32
ip prefix-list EIGRP_INTO_OSPF seq 15 permit 192.168.26.32/29 le 32
ip prefix-list EIGRP_INTO_OSPF seq 20 permit 192.168.30.72/29 le 32
ip prefix-list EIGRP_INTO_OSPF seq 25 permit 192.168.20.128/25 le 32
ip prefix-list EIGRP_INTO_OSPF seq 26 permit 192.168.101.0/24 le 32 <- Backup Route for MPLS Remote Office
route-map EIGRP_INTO_OSPF permit 10
match ip address prefix-list EIGRP_INTO_OSPF

So in the case of a /24. If it were say broken up into /25's? From our remote sites we are using aggregate-address summary-only. Not sure how I would advertise a more specific route via BGP, sorry.
I didnt have this problem until I moved my firewalls. They plugged into the cores where IBGP was running and the static never kicked in unless the bgp route disappeared. I guess I could use my static redistribution for my VPN sites and use statics across the cores for the handful of backup links I have.

Real-time traffic, static routes, ports, filters, EIGRP.

I am trying to send realtime traffic to a host device using a specific port on that end device. The path between the two devices currently consists of two EIGRP gigabit hops.
I will be adding a single hop 100Mb link between the originating router and the destination router. I want only the specific host to host realtime traffic to cross this 100Mb link, ie. no EIGRP traffic should cross it.
Using this 100Mb link, if you set up a static route from the source router to the destination router then you can't control who uses that route ie 100Mb link. If you use EIGRP and do a route filter, you end up with another problem, EIGRP traffic crossing the link.
Suggestions?

Bruce
If I understand your requirements correctly then the solution you are looking for is Policy Based Routing. PBR can alter (or over-ride) the routing decision that would be made in the routing table. You configure PBR with a route map. Frequently the route map uses a access list to identify the traffic to be treated differently. So you could create an access list that would identify the specific host to host traffic and then in the route map you can set the next hop to use the special link.
In this solution you would not run any routing protocol over the special link. Only the Policy Routed traffic would use the link.
HTH
Rick

Memory allocation when using multiple instances of a static array.

Hi,
I have question regarding the amount of memory allocated using multiple instances of a class that has declared a static array, e.g.:
package staticTest;
import javacard.framework.*;
public class StaticClass
     static byte[] staticArray;
     public StaticClass()
          staticArray = new byte[100];
/* Another class */
package statictest;
import javacard.framework.*;
public class MyApplet extends javacard.framework.Applet
     StaticClass staticLibA, staticLibB;
     public MyApplet()
          /* Here we use two instances */
          staticLibA = new StaticClass();
          staticLibB = new StaticClass();
     public static void install(byte[] bArray, short bOffset, byte bLength)
          (new MyApplet()).register(bArray, (short)(bOffset + 1), bArray[bOffset]);
     public void process(APDU apdu)
          byte[] buf = apdu.getBuffer();
          switch(buf[ISO7816.OFFSET_INS])
}When downloading and installing this applet on a gemXpresso 211 PKis (that has a ' Get free EEPROM memory' function) the card reports that the amount of memory is increased by a approx 100 bytes every time a new staticLibX is added in the MyApplet constructor. Is this correct? If yes, can someone explain to me how to declare a static array that only allocates memory once and are being shared by the instances.
Best regards,
Jonas N

Hi!
I have another issue about static variable. The following is my sample code:
========================================
package com.Test01;
import javacard.framework.*;
import javacard.security.*;
import javacardx.crypto.Cipher;
public class Test01 extends Applet {
     OwnerPIN Pin;
     static DESKey[] keys;
protected Test01(byte[] buffer, short offset, byte length) {
keys = new DESKey[4];
length = 0;
while (length < 4) {
     keys[length] = (DESKey)KeyBuilder.buildKey((byte)3, (short)0x80, false);
     length = (byte)(length + 1);
public static void install(byte buffer[], short offset, byte length) {
new Test01(buffer, offset, length);
===========================================================
If there are two instances, A and B, created in the package.
My issue:
1. Are keys[0]~ keys [3] kept while B is deleted?
2. Does each instance have itsown object while "keys[length] = (DESKey)KeyBuilder.buildKey((byte)3, (short)0x80, false);"?
3. follow item 2, if A and B share the same object, is the object kept while B is deleted?
Thank you
Best regards,
kantie

Default static route and Null 0

Hi Everyone,
Need to clear some doubts for below setup
Switch 3550A is connected to Internet Router and has OSPF nei relationship with it.
3550A#                      sh run int fa0/11
Building configuration...
Current configuration : 272 bytes
interface FastEthernet0/11
description OSPF LAN Connection to 2691 Router Interface Fas 0/1
no switchport
ip address 192.168.5.2 255.255.255.254
sh ip route shows
3550A#sh ip route
Gateway of last resort is 192.168.5.3 to network 0.0.0.0
O*E2 0.0.0.0/0 [110/1] via 192.168.5.3, 20:39:56, FastEthernet0/11
3550A#
All is working fine.
For testing purposes i config below static route on 3550A
ip default-network 192.168.1.0
ip route 192.168.1.0 255.255.255.0 Null0
After above change
3550A#           sh ip route
Codes: C - connected, S - static, R - RIP, M - mobile, B - BGP
       D - EIGRP, EX - EIGRP external, O - OSPF, IA - OSPF inter area
       N1 - OSPF NSSA external type 1, N2 - OSPF NSSA external type 2
       E1 - OSPF external type 1, E2 - OSPF external type 2
       i - IS-IS, su - IS-IS summary, L1 - IS-IS level-1, L2 - IS-IS level-2
       ia - IS-IS inter area, * - candidate default, U - per-user static route
       o - ODR, P - periodic downloaded static route
Gateway of last resort is not set
S*   192.168.1.0/24 is directly connected, Null0
O*E2 0.0.0.0/0 [110/1] via 192.168.5.3, 20:38:38, FastEthernet0/11
Now i can not ping to internet as below
3550A#ping 4.2.2.2
Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 4.2.2.2, timeout is 2 seconds:
Success rate is 0 percent (0/5)
When we ping from Switch then source IP is always the Outside interface IP right?
So in this case Switch is using which IP as source?
Ping to internet is not working as default network is set to 192.168.1.0 and all request goes to this IP and then it goes to
Null interface right?
Extended ping works fine as below
3550A#ping
Protocol [ip]:
Target IP address: 4.2.2.2
Repeat count [5]:
Datagram size [100]:
Timeout in seconds [2]:
Extended commands [n]: y
Source address or interface: 192.168.5.2
Type of service [0]:
Set DF bit in IP header? [no]:
Validate reply data? [no]:
Data pattern [0xABCD]:
Loose, Strict, Record, Timestamp, Verbose[none]:
Sweep range of sizes [n]:
Type escape sequence to abort.
Sending 5, 100-byte ICMP Echos to 4.2.2.2, timeout is 2 seconds:
Packet sent with a source address of 192.168.5.2
Success rate is 100 percent (5/5), round-trip min/avg/max = 76/79/80 ms
Second thing to confirm is this ping works because 192.168.5.2 is directly connected to Internet Router interface?
Regards
MAhesh

Hi Mahesh,
When we ping from Switch then source IP is always the Outside interface IP right?
That is correct. By default it is always the outgoing interface on the device unless you specify it differently.
Ping to internet is not working as default network is set to 192.168.1.0 and all request goes to this IP and then it goes to
Null interface right?
That is correct. Null0 can't be used as next-hop.
Second thing to confirm is this ping works because 192.168.5.2 is directly connected to Internet Router interface?
No, that is because 192.168.5.0/30 is NATed. Remember 192.168.x.x address is a private segment and cannot access the Internet unless NAT is used.
HTH
Reza

How to reference multiple instances of the same Java object from PL/SQL?

Dear all,
I'm experimenting with calling Java from PL/SQL.
My simple attempts work, which is calling public static [java] methods through PL/SQL wrappers from SQL (and PL/SQL). (See my example code below).
However it is the limitation of the public static methods that puzzels me.
I would like to do the following:
- from PL/SQL (in essence it needs to become a forms app) create one or more objects in the java realm
- from PL/SQL alter properties of a java object
- from PL/SQL call methods on a java object
However I fail to see how I can create multiple instances of an object and reference one particular object in the java realm through public static methods.
My current solution is the singleton pattern: of said java object I have only 1 copy, so I do not need to know a reference to it.
I can just assume that there will only ever be 1 of said object.
But I should be able to make more then 1 instance of an object.
To make it more specific:
- suppose I have the object car in the java realm
- from PL/SQL I want to create a car in the java realm
- from PL/SQL I need to give it license plates
- I need to start the engine of a scpecific car
However if I want more then 1 car then I need to be able to refrence them. How is this done?
Somehow I need to be able to execute the following in PL/SQL:
DECLARE
vMyCar_Porsche CAR;
vMyCar_Fiat CAR;
BEGIN
vMyCar_Porsche = new CAR();
vMyCar_Fiat = new CAR();
vMyCar_Porsche.setLicensePlates('FAST');
vMyCar_Porsche.startEngine();
vMyCar_Fiat.killEngine();
END;
Thanks in advance.
Best Regards,
Ruben
My current example code is the following:
JAVA:
===
CREATE OR REPLACE AND RESOLVE JAVA SOURCE NAMED CODAROUL."RMG/BO/RMG_OBJECT" as package RMG.BO;
public class RMG_OBJECT {
private static RMG_OBJECT instance = new RMGOBJECT();
private String rmgObjectNaam;
private RMG_OBJECT(){
this.rmgObjectNaam = "NonDetermined";
public static String GET_RMGOBJECT_NAAM () {
String toestand = null;
if (_instance == null) {toestand = "DOES NOT EXIST";} else { toestand = "EXISTS";};
System.out.println("instance : " + toestand);
System.out.println("object name is : " + _instance.rmgObjectNaam);
return _instance.rmgObjectNaam;
public static Integer SET_RMGOBJECT_NAAM (String IN)
try
_instance.rmgObjectNaam = IN;
return 1;
catch (Exception e)//catch
System.out.println("Other Exception: " + e.toString());
e.printStackTrace();
return 5;
} //catch
PL/SQL Wrapper:
==========
CREATE OR REPLACE FUNCTION CODAROUL.SET_RMGOBJECT_NAAM(NAAM IN VARCHAR2) return NUMBER AS
LANGUAGE JAVA NAME 'RMG.BO.RMG_OBJECT.SET_RMGOBJECT_NAAM (java.lang.String) return java.lang.Integer';
Calling from SQL:
==========
CALL dbms_java.set_output(2000);
select CODAROUL.GET_RMGOBJECT_NAAM() from dual;
Edited by: RubenS_BE on Apr 6, 2012 5:35 AM
Edited by: 925945 on Apr 6, 2012 5:41 AM

You can do this by manually creating a new iterator binding in your binding tab.
So instead of dragging the VO directly to the page, go to the binding tab, add a new executable iterator binding, and point to that one from your ELs in the page itself.

Opening multiple instances of the same vi

Hi,
I am having trouble opening multiple instances of the same vi.
This vi that I am trying to create can be used like a message display window to display numbers, strings etc (just like labVIEW's display message control except that this vi will update the displayed controls every scan). So for example, I will like to have a subvi, lets call it popup.vi, where I wire in three inputs from my main vi (2 strings and 1 number). Now in my main.vi I have 2 numeric controls Num1 and Num2 with their KeyFocus property nodes.
The intention is that when I select Num1 or Num2 numeric control with my mouse the KeyFocus boolean becomes true and the popup.vi logic kicks in. Meaning that the popup.vi runs and opens its front pane
l displaying the two strings and one numeric value from the main.vi.
The problem arises when I want to open and run multiple instances of the popup.vi.
So when I click on Num1 with my mouse I want the popup.vi to open its front panel with a set of strings and number displayed. At the same time I would want to click on Num2 control and open another instance of the popup.vi with another set of strings and numeric value.
Also, I saved the popup.vi as popup.vit but for some reason vi server will not open the popup.vit. I get a message that popup.vit is already in memory - cannot load.
I have attached a sample of the main.vi and popup.vi program.
Please let me know if you need further clarifications.
I appreciate your help in this matter.
Thanks
Nish
Attachments:
popupVI.zip ‏37 KB

You have to clone the VI. You say you have tried to use the .vit way of cloning...but in the example there is no use of VIT and the VI is called statically. If you want to open multiple instances of the same VI and the VI has a user interface (or needs to be a unique instance of the VI every time even though it's called the same place in the diagram, not the case here though) you need to do it dynamically, i.e. with an invoke node.
There is actually a problem doing just that, it causes a memory leak...but that won't be a problem unless the VI is called frequently / the software will be running over very long periods of time. You can see an example of it here (attached to my reply to the question):
http://exchange.ni.com/servlet/ProcessRequest?RHIVEID=101&
RPAGEID=135&HOID=506500000008000000A07D0000&UCATEGORY_0=_49_%24_6_&UCATEGORY_S=0&USEARCHCONTEXT_QUESTION_0=VI+cloning&USEARCHCONTEXT_QUESTION_S=0
About the memory leak problem:
http://exchange.ni.com/servlet/ProcessRequest?RHIVEID=101&RPAGEID=135&HOID=506500000008000000B7A40000&UCATEGORY_0=_49_%24_6_&UCATEGORY_S=0&USEARCHCONTEXT_QUESTION_0=VI+cloning&USEARCHCONTEXT_QUESTION_S=0
MTO

Multiple instances of FileOutputStream

Hi, I'm learning how to use Java's multithreading features. I created a class called "ObjectMaker". An instance of this class serializes some objects into a file, using a FileOutputStream (inside an ObjectOutputStream) for writing. In the mean time an other object of another class deserializes those objects reading from the file. Each concurrent access to the file (for writing,reading) is synchronized. Here's the problem:[ u]if each writer has its own FileOutputStream, then the concurrent writing on the file doesn't work (despite of every writer is synchronized on the same object), and the reader aren't able to read anything from the file; if the FileOutputStream is static (every instance share the same FileOutputStream reference) everything works. I read on Documentation that some OS "allow a file to be opened for writing by only one FileOutputStream (or other file-writing object) at a time" .But it also says that I would recive an exception from FileOutputStream if I try to open a file for writing that is already open by another FileOutputStream....and It doesn't happen, no exception are thrown.Then here is the problem:It's possible to manage Multiple instances of FileOutputStream linked on the same file?

Hello Joseph,
Have u been successful on the installation.
There were no replies after yours.
Can you please let me know about installing 2 MDS instances on the same host.
If you have any document you followed please mail me to [email protected]
It will be more helpful as the dead line to this is killing me.
Regards,
VishnuM

Multiple vlans configuration issue with RV016 router and SG 300-10MP witch

Hi,
I have to configure multiple vlans served with a unique DCHP server . As first step, I just will The DHCP server to serve 2 vlans. The following is the hardware and configuration that I implemented :
Router (RV016 10/100 16-Port VPN Router) as gateway mode:
IP : 172.16.0.1/24
DHCP Server :
IP : 172.16.0.2/24 GW: 172.16.0.1
2 subnets :
172.16.1.0/24 GW: 172.16.1.1 to serve vlan 1
172.16.2.0/24 GW:172.16.2.1 to serve vlan 2
Switch (SG 300-10MP 10-Port Gigabit PoE Managed Switch) as layer 3 mode:
IP 172.16.0.254 (vlan 8 default)
Vlan 1 : 172.16.1.1
Vlan 2 : 172.16.2.1
1 device connected on each vlan
a workstation on the vlan 1
a laptop on the vlan 2
In this scenario (see the attached pdf file) the DHCP server is connected on a router, hosts on vlans dont receive any IP address.
But If I connect the DHCP server on a trunked switch port and adapt the DHCP server gateway 172.16.0.1 to 172.16.0.254, hosts receive ip address properly.
I have to connect the DHCP server directly to the router. How can I do that, what is wrong in the configuration ?
I hope the explanations are clear enough and my English too
Any help will be highly appreciated,
Zoubeir

Hi Eric, the small business group doesn't support the ASA config, but I can help with the switch.
A couple things I notice in your description-
48 port (192.168.1.254) and the other 24P (192.168.1.253) we have a second vlan 20 set up on the 24P switch (192.168.2.253) we have ports 1-12 set for vlan20 (untagged and trunk), the remaining ports on on the default vlan 1.
The connection between the switches, is it 1u, 2t?
The link between the switches should be 1u, 2t, the switches support the trunking and vlan tagging, meaning all communication will work fine.
We have the 24p and 48p switches connect using GE1 and GE1. We are unable to ping a device on vlan 20 ( on the 24p switch
The 24p switch should be in layer 2 mode, if you have the 48 port l3 switch upstream. Additionally, you need to have the default gateway set on the 24p switch.
We have a static route set on the 24p switch (0.0.0.0 192.168.1.0).
Between the switches, it shouldn't require any static routes, assuming you correctly trunk / tag your ge1 ports, with both switches operating in l3, the ip route table dynamically builds the connected routes, therefore a static route is redundant.
-Tom
Please rate helpful posts

Sql server with multiple instance

hi All,
i want to install sql server with two instance and i want to assign two separate is ips for both default instance and named instance and both should work like
two different server
note: If i logged with one ip(default instance) it should show its own databases not both or named instance databases
thanks
chintala

Hi chintala,
According to your description, if you want to assign the two different IP to stand for default instance
and named instance, I recommend you set up aliases on a SQL Server with multiple instances. For example, a physical server is named WinServer and 2 instances of SQL Server named WinServer and WinServer\instance1.
We need to add 1 IP addresse to the WinServer, there is one IP address already assigned to WinServer, 192.168.1.1, by adding one new IP's, 192.168.1.2 .
We need to move into DNS to configure our host names and aliases. The physical machine must be steup as a static IP in DNS. SQLAlias1, which point to the default instance on WinServer, will be setup as a DNS with an address of 192.168.1.1.
SQLAlias2, which will point to WinServer\Instance1, will be setup as a new Host (A) record in DNS with an address of 192.168.1.2.
When everything has been completed on the DNS side, you need to open up SQL Server Configuration Manager on WinServer and configure the protocols for each instance one at a time, restart the 3 SQL services.
For more information about how to set up aliases on a SQL Server with multiple instances, you can review the similar article.
http://kevine323.blogspot.com/2013/08/setting-up-aliases-on-sql-server-with.html
Regards,
Sofiya Li
Sofiya Li
TechNet Community Support

Cisco UC560 Not Clearing Static Routes When VPN Connections Drop

We have a Cisco UC560 (UC560-FXO-K9) running "Cisco IOS Software, UC500 Software (UC500-ADVIPSERVICESK9-M),
Version 15.1(2)T2, RELEASE SOFTWARE (fc1)" The issue is when we have end users connecting with the Cisco VPN Client to this device sometimes we are unable to connect to any devices on our LAN or sometimes we can't connect to the LAN on the other end of our site-to-site VPN. The one symptom I've observed when this happens is that old VPN sessions that have disconnected appear to leave static routes from the user's outside IP at their home to an IP on our LAN to a Virtual-Access interface. When this starts to happen, I restart the firewall to clear out the stale static routes and the problem is fixed, for a while at least. Below is the current state where we have the site-to-site VPN connected to our branch office and 2 user's connected with Cisco VPN clients. Below that is the static route table which has 5 total Virtual-Access interface routes (one is an extra route for a user currently connected so that their outside IP is in the static route table with 2 inside IP's associated.) Is there a way to fix the cleanup of VPN connections when they terminate?
#sh crypto isakmp peers
Peer: <branch office outside IP> Port: 500 Local: <firewall's outside IP>
Phase1 id: <branch office outside IP>
Peer: <users's outside IP #1> Port: 50420 Local: <firewall's outside IP>
Phase1 id: EZVPN_GRP_437
Peer: <user's outside IP #2> Port: 49345 Local: <firewall's outside IP>
Phase1 id: EZVPN_GRP_437
Bugsy#sh ip ro st
Codes: L - local, C - connected, S - static, R - RIP, M - mobile, B - BGP
       D - EIGRP, EX - EIGRP external, O - OSPF, IA - OSPF inter area
       N1 - OSPF NSSA external type 1, N2 - OSPF NSSA external type 2
       E1 - OSPF external type 1, E2 - OSPF external type 2
       i - IS-IS, su - IS-IS summary, L1 - IS-IS level-1, L2 - IS-IS level-2
       ia - IS-IS inter area, * - candidate default, U - per-user static route
       o - ODR, P - periodic downloaded static route, H - NHRP, l - LISP
       + - replicated route, % - next hop override
Gateway of last resort is <next hop of ISP for firewall> to network 0.0.0.0
S*    0.0.0.0/0 [1/0] via <next hop of ISP for firewall>
      10.0.0.0/8 is variably subnetted, 12 subnets, 3 masks
S        10.0.0.153/32 [1/0] via <non-connected IP of VPN user>, Virtual-Access2
S        10.0.0.155/32 [1/0] via <non-connected IP of VPN user>, Virtual-Access2
S        10.0.0.156/32 [1/0] via <user's outside IP #2>, Virtual-Access3
S        10.0.0.158/32 [1/0] via <user's outside IP #1>, Virtual-Access3
S        10.0.0.159/32 [1/0] via <user's outside IP #2 again>, Virtual-Access2
S        10.1.10.1/32 is directly connected, Vlan90

Hi Brian,
This sounds like you are running into the following known issue:
CSCtl03682 - EzVPN client: Several RRI routes pointing to same virtual interface
which is Dup'd to:
CSCtf39056 - RRI routes not deleted
This is fixed since 15.1(2)T4, so I would recommend upgrading to SWP 8.2 or higher. The only other way to clean up the stuck routes is to reload the router.
Thanks,
Brandon

Loading multiple instance of a

Hi,
I'm trying to develop a Java application with Avetana Bluetooth. This library is a complete stack for Bluetooth in Java. The big problem is the hardware interface: Avetana can get only one instance for each java application running on it. So, if I have multiple hardware interface on my system, I must run multiple JVM with my application...
This happen because the Avetana library have a lot of static variable to hold some reference to the bluetooth hardware.
The big question is: can I make some Java code to create & run multiple instance of my Bluetotth code, so each Avetana Library is TRULY loaded in a ISOLATED context (so the static variables are independent each from others)?
I've tryed with custom classloading but nothing...
Here is the code from main() on StaticTestClassLoader (a code used for testing).
MyBT is the Bluetooth class wich load the Avetana library on hardware interface (specified by the method initStack(int hardwareInterfaceID)).
URL[] url = { new File("subdir").toURL() }; // because MyBT.class is in the "subdir" folder
URLClassLoader cl1 = new URLClassLoader(url);
URLClassLoader cl2 = new URLClassLoader(url);
URLClassLoader cl3 = new URLClassLoader(url);
Class<MyBT> c1 = (Class<Dummy>)cl1.loadClass("MyBT");
Class<MyBT> c2 = (Class<Dummy>)cl2.loadClass("MyBT");
Class<MyBT> c3 = (Class<Dummy>)cl3.loadClass("MyBT");
Object bt1 = c1.newInstance(); // this is line 22 where exception rise up
Object bt2 = c2.newInstance();
Object bt3 = c3.newInstance();
((MyBT)bt1).initStack(0); // run my Bluetooth application on hardware interface #1
((MyBT)bt1).initStack(1); // the same but on hardware interface #2
((MyBT)bt1).initStack(2); // the same but on hardware interface #3No error while compiling, but when running:
Exception in thread "main" java.lang.NoClassDefFoundError: MyBT
at StaticTestClassLoader.main(StaticTestClassLoader.java:22)this happen if I didn't add on classpath the MyBT class in the "subdir". If I add it, it works but all three instance are in the same context and the library doen't load the right hardware interface...
If I omit the obejct assignement with bt1, bt2 and bt3 it works in both scenarios, but I haven't any reference to the new instantiated classes ...(so I can't call methods on there!)
Any idea, web link or suggestion?
Thanks in advance, best regards,
Francesco Pessina.

I think the question is detalied in an unclear manner. The substance of my question is:
- I have some code in a library with static variables.
- if I load (instantiate) this library one, twice or more, this static variables have all the same value because are static and live in the same JVM/context
- the need is to load & use this library multiple time (two or more "instance" of it) from the same application in a single JVM.
So my trouble is to write some Java code that create a runtime different context spaces in the (single) JVM so I can run this "static" library more times (one library for each context).
I've tryed with classloader and singleton but all instantiated class have always the same context, so the library have (in all the created instance) always the same static variable values.
Sorry for my english, have a good new year!
Regards,
Francesco.

Multiple instance of custom attributes ACS 5.x

Hello,
is there a way to have multiple instances of user custom attributes and
insert those as multiple instances of the A/V Pair in the authorisation profile in ACS 5.2/5.3 ?
Background:
We have to migrate a ACS 4.2 to 5.3.
In ACS 4.2 our client used the multiline attribute
Number
Name
Description
Type of Value
Inbound/Outbound
Multiple
22
Framed-Route
Routing information to configure for the user on this AAA client. The RADIUS RFC (Request for Comments) format (net/bits [router [metric]]) and the old style dotted mask (net mask [router [metric]]) are supported. If the router field is omitted or zero (0), the peer IP address is used. Metrics are ignored.
String
Outbound
Yes
to specify multiple routes to various networks in the RADIUS reply spcific for every single PPP username of routers dialing in.
Using the internal user database, extended by a string attribute and using that attribute as source of a dynamic value
in the access-policy works basically.
But as I have only ONE single line instance of the attribute for every user, I can only return ONE framed-route.
We have lots of cases where multiple routes have to be assigned to one router.
I 'd like to avoid defining a seperate access profile for every remote RAS router for external PPP Dial-In...
I Think Jack here https://supportforums.cisco.com/thread/2032506 has a simmilar issue...
Any idea?
Thanks, Frank

I had to do this once and I created several custom attributes
-Route1 (String)
-Route2 (String)
-Route3 (String)
etc ...
And in the authorization profile you return all those custom attributes as Framed-route. I was actually sending up to 10 Framed-Routes contained in 10 different attributes.

Connecting MDM to multiple instances of XI

Hi All,
does anyone know if it is a) possible and b) sensible to connect a single instance of MDM to multiple instances of XI?
We have two regional XI instances set-up, but a single MDM instance being planned - I guess the question I'm asking is, can we use the single MDM instance to manage the data for all the client systems we have attached to both instances of XI??? (e.g. region 1 R/3, region 2 R/3, region 1 BW, region 2 BW etc).
Thanks in advance for any help you can give me!
Cheers
Tim.

Hi Tim,
I guess this is possible. If you set-up syndication, you will need to ensure you use different ports/folders for each of the XI systems and then configure XI to fetch the files from the folder assigned to them.
I think you could also verify if one XI could be the primary one, routing information to the other box if needed.
Regards,
Dirk

Multiple instances of EIGRP or static routes

Similar Messages

Maybe you are looking for