WAAS + Copying Windows files = Error

Unanswered Question

Hello all,

I am trying to setup an optimized Datacenter/Branch WAAS solution.  In my datacenter, I have my CM connected to the user VLAN.  I have my core-wae (512) connected to my WAN router (2811) on a dedicated ethernet interface.  At my branch, I have my edge-wae (also a 512) connected to a dedicated ethernet interface on the branch WAN router (1841).  For both routers, I have WCCPv2 enabled with "IP WCCP 61 redirect in" configured on the LAN facing interface and "IP WCCP 62 redirect in" on the WAN facing interface.  From a computer at the branch office, I can browse windows based shares.  When I try to copy a file from a server (at the datacenter) to the computer, It hangs and eventually crashes.  If I attempt to copy from a command windows, copying the same file, after a bit of time, I get the error "The network service is no longer available" (or something like that) error message and nothing gets copied.  If I disable WCCP on the routers, I can copy the files without error.

Any ideas would be greatly appreciated.

Eric

I have this problem too.
0 votes
  • 1
  • 2
  • 3
  • 4
  • 5
Overall Rating: 0 (0 ratings)
Loading.
Zach Seils Mon, 03/01/2010 - 14:22

Hi Eric,

A few questions:

  • Is other types of (TCP) traffic to pass between the branch and Data Center when WCCP is enabled?
  • Do you have the Enteprise license and the CIFS AO enabled?
  • How is the connection between the client and server being handled (sh stat conn)?
  • What OS/version is being used on the client and server?
  • Do you see any suspicious messages in the syslog.txt file on either WAAS device?

Thanks,

Zach

Zach,

Thanks for the quick response...

A few answers:

  • Is other types of (TCP) traffic to pass between the branch and Data Center when WCCP is enabled?

Yes.   Each WAE is pretty much stock configured

  • Do you have the Enteprise license and the CIFS AO enabled?

Yes, The CIFS Application Optimizer is enabled.

  • How is the connection between the client and server being handled (sh stat conn)?

Currently, All ConnTypes = PT No Peer.  I have WCCP disabled because of the errors.

  • What OS/version is being used on the client and server?

The client computer is a Windows XP sp3 computer and the server is a Windows 2003 STND server

  • Do you see any suspicious messages in the syslog.txt file on either WAAS device?

I do see a bunch of "Routing Loop Detected - Same devid - packet dropped

I don't think I have any routing loops in between my branch and datacenter (puzzled).

Thanks in advance,

Eric

Zach Seils Tue, 03/02/2010 - 05:49

What are the IP addresses and TCP ports in the routing loop messages?  Are they the same as the client/server you are performing the copy

between?  A good way to start troubleshooting why these messages are happening is:

  1. Note the source/destination IP addresses in the routing loop messages.
  2. From the WCCP-enabled router, check the routing table (sh ip ro) for the destination IP address.
  3. Go to the next-hop in the output of (2) above and check the routing table for the destination.

If you'd like to provide the router and WAE configurations, as well as the output from sh ip ro on each router, I can take a look as well.

Regards,

Zach

Zach,

Thanks for replay.

The source and destination IP and ports shown in the syslog.txt on branch WCCP enabled router are from various clients in the branch office including the one that I was testing from.  The WCCP enabled branch router has a multilink interface made up of 2xT1 circuits.  This multilink interface connects to a hosted MPLS cloud.  My source IP address of the client is 10.6.3.7 (in the branch office).  The server (in the datacenter) is 10.1.3.55.  On the branch office WCCP enabled router, "show ip route 10.1.3.55" comes back with "Subnet not in table".  This would indicate that the route for 10.1.3.55 goes through the default route - through the Multilink1 interface.

Gateway of last resort is 0.0.0.0 to network 0.0.0.0

216.149.520/24 is variably subnetted, 2 subnets, 2 masks

C     216.149.52.109/32 is directly connected, Multilink1

C     216.149.52.108/30 is directly connected, Multilink1

10.0.0.0/8 is variably subnetted, 3 subnets, 3 masks

C     10.6.0.0/16 is directly connected, FastEthernet0/0

C    10.50.6.0/24 is directly connected, FastEthernet0/1

C    10.253.6.1/32 is directly connected, Loopback0

S*     0.0.0.0/0 is direclty connected, Multlink1

Eric

Zach Seils Tue, 03/02/2010 - 10:15

Yes, the router on the other end of the multilink interface.  I basically want you to trace the routing hop-by-hop until we find where it is looping.

Zach

What are the chances that the routing loop gets created when I enable WCCP between the branch wae and the branch router.  The branch wae is directly connected via cross-over to a dedicated ethernet port on the branch router.  I do have the "ip wccp redirect exclude in" command on the wae-dedicated interferace on the router.

Zach Seils Tue, 03/02/2010 - 13:26

This shouldn't be a problem.  Let's try the traceroute and see what the results are.

Zach

Zach Seils Tue, 03/02/2010 - 13:25

Somehow traffic destined to the IP address/port in the syslog messages is being seen twice by WAAS.  Can you perform a traceroute sourced from and destined to the IP addresses in the syslog message?

Zach

Zach,

Here is a series of traceroutes orginating from the workstation.  I don't have access the 216.149.x.x routers as they are located in the MPLS cloud.

C:\>tracert 10.1.3.55

Tracing route to acct-old.lbbslaw.com [10.1.3.55]
over a maximum of 30 hops:

  1    <1 ms    <1 ms    <1 ms  10.6.1.1
  2    <1 ms    <1 ms     1 ms  10.6.128.254
  3    13 ms    28 ms    25 ms  216.149.52.109.ptr.us.xo.net [216.149.52.109]
  4    16 ms    27 ms    32 ms  216.149.96.201.ptr.us.xo.net [216.149.96.201]
  5    20 ms    15 ms    32 ms  216.149.33.102.ptr.us.xo.net [216.149.33.102]
  6    16 ms    17 ms    24 ms  acct-old.lbbslaw.com [10.1.3.55]

Trace complete.


LDBB3825-SF#traceroute ip 10.1.3.55 source 10.6.1.1

Type escape sequence to abort.
Tracing the route to 10.1.3.55

  1 10.6.128.254 0 msec 0 msec 0 msec
  2 216.149.52.109.ptr.us.xo.net (216.149.52.109) 4 msec 4 msec 4 msec
  3 216.149.96.201.ptr.us.xo.net (216.149.96.201) 16 msec 16 msec 16 msec
  4 216.149.33.102.ptr.us.xo.net (216.149.33.102) 16 msec 16 msec 20 msec
  5 10.1.3.55 32 msec 16 msec 16 msec
LDBB3825-SF#


lbbs_SanFranCa_ipvpn#traceroute ip 10.1.3.55 source 10.6.128.254

Type escape sequence to abort.
Tracing the route to 10.1.3.55

  1 216.149.52.109 8 msec 4 msec 4 msec
  2 216.149.96.201 16 msec 16 msec 16 msec
  3 216.149.33.102 16 msec 16 msec 20 msec
  4 10.1.3.55 16 msec 16 msec 16 msec
lbbs_SanFranCa_ipvpn#

Zach Seils Wed, 03/03/2010 - 08:08

Eric,

Thanks for the information.  In lacorertr.cfg, I don't see service group 61 applied to any interfaces.  Other than that, I don't see anything wrong with the configuration or routing.  Would you be willing to do a WebEx so we can troubleshoot live?

Zach

Jan Rockstedt Wed, 03/03/2010 - 01:58

Eric,

Do you have Digitally sign enable on the server?

See below.

We have seen a lot of issues with this.

Example:

Not able to mount remote drive.

Security problem on folders, solved when we remapped the drive.

Off and on problem to copy file to remoted map drives from diffrent program och from windows explorer.

Jan Rockstedt

Actions

This Discussion