ANR8213W Session open with node name timed out.

[RonJ]

Hello all,



I have a new server with AIX Tivoli 5.3 attempting to run a backup schedule for a windows workstation with 5.3 client. I am able to run manual backups, and the Scheduler service is started. I have insured the systems have the correct DNS names and what the server thinks the client is named. Have double checked IP addresses and ports. Anybody have any ideas? Thanks, RonJ





ANR8213W Session open with node name timed out.

Explanation: The ADSM server attempts to contact the indicated node name. The attempt is unsuccessful.

System Action: The attempt to communicate with the indicated node fails.

User Response: Ensure that the node name is properly specified and that it is accessible over the network. Use the TCP/IP ping command for this purpose.

 

[sgabriel62]

Questions

Firewalls rules in place?

What are your network timeouts?

Can you execute the dsmc command and contact the server in a reasonable amount of time?

Are you polling or is the schedule prompting the client?

 

[mikey65]

Although the IBM technote isn't exactly the same question, you might find the toubleshooting portions very helpful...I did. The article explains how to resolve some communication problems.



IBM Technote



The following is an excert from the article:

tsm: SERVER1> create sqlt Nodes mynodes

tsm: SERVER1> create sqlt Schedule.Node.Addresses myaddr

tsm: SERVER1> select mynodes.c0 as "Node name",myaddr.c2 as "Contact address" from mynodes,myaddr where myaddr.c0=mynodes.c1

tsm: SERVER1> drop sqlt mynodes

tsm: SERVER1> drop sqlt myaddr



This will produce a list in two columns, the first showing the nodename and the second showing the IP address that the TSM scheduler will use to contact the node for a prompted schedule.





I hope this helps,

mike

 

[ironfang]

I understand this is an old thread but it is relevant to me. I have been having problems with a client not backing up. I traced the error to this post and when I run this command

tsm: SERVER1> select mynodes.c0 as "Node name",myaddr.c2 as "Contact address" from mynodes,myaddr where myaddr.c0=mynodes.c1

My first entry the one I am having problems with is not correct. It has the NETBIOS name of our domain as the address. I have tried upating HLADDRESS, I have completly removed the node including filespace and recreated it and it still comes back. If I do a "q node nodename f=d" everything shows as correct.

Any ideas how I may be able to fix this?

Thanks. First day having to take over TSM withi nothing but Google and you guys and I think I am already in over my head.

 

[Harry_Redl]

Hi,

just to confirm:
a) when trying to "ping" the address of the node from the server - does it resolve correctly? Any possible problems with /etc/hosts?
b) seems you are using PROMPTED scheduler mode - can you "telnet <node_address> 1581" from the server? Does the client listen?
c) have you tried IP (not hostname, FQDN ...) as the HLADDRESS?
d) what about POLLING mode (so the client contacts the server) - is it an option?

Harry

 

[ironfang]

Yes I can resolve the name correctly through nslookup or ping. I have tried to update HLADDRESS as static IP. Telnet does not connect at 1581 or 1500 but does connect at 1501.

I could probably use those other methods I just do not know how yet.

I am wondering where it is picking up that odd name from and why it will not go away.

 

[Harry_Redl]

Hi,

can I see the "q node f=d" output and the dsm.sys/dsm.opt from the client?

Thanks

Harry

 

[LED888]

On the system that is hosting the TSM Client, what messages doe we see in the dsmerror.log and dsmsched.log when the ANR8213W message is produce?

In the actlog what messages are we seeing leading up to
"ANR8213W Session open with node name timed out".

On the TSM Server what is the idletimeout and the commtimeout set to?
It could be due to either the idletimeout or the commtimeout set too low or not high enough.

Good Luck,
Sias

 

[ironfang]

I was getting ready to post that deleting the filespace and removing the node and recreating it did not fix the problem. However recreating it with a new name did. Until last night. This node now has the wrong contact info again and is trying to contact INFO_TECH which is our PRE2000 Domain Name.


I also have another node that seems to be doing the same thing with a different name. It is off-site and I do not know where this new name comes from.




tsm: TSM_SERVER1>q node appsrv2008 f=d


Node Name: APPSRV2008
Platform: WinNT
Client OS Level: 6.01
Client Version: Version 6, Release 2, Level 2.0
Policy Domain Name: ADMIN-STANDARD
Last Access Date/Time: 09/19/2011 23:51:29
Days Since Last Access: 2
Password Set Date/Time: 09/14/2011 15:05:19
Days Since Password Set: 7
Invalid Sign-on Count: 0
Locked?: No
Contact:
Compression: Client
Archive Delete Allowed?: Yes
Backup Delete Allowed?: No
Registration Date/Time: 09/14/2011 15:05:19
Registering Administrator: ADMIN
Last Communication Method Used: Tcp/Ip
Bytes Received Last Session: 4,719.99 M
Bytes Sent Last Session: 124.86 M
Duration of Last Session: 2,461.00
Pct. Idle Wait Last Session: 24.38
more... (<ENTER> to continue, 'C' to cancel)


Pct. Comm. Wait Last Session: 60.22
Pct. Media Wait Last Session: 0.00
Optionset: WINDOWS
URL: http://myclient.mycompany.com:1581
Node Type: Client
Password Expiration Period: 0 Day(s)
Keep Mount Point?: No
Maximum Mount Points Allowed: 1
Auto Filespace Rename : No
Validate Protocol: No
TCP/IP Name: APPSERVER
TCP/IP Address: 192.168.1.131
Globally Unique ID: 9a.38.77.40.18.35.11.e0.bc.b3.e4.1
Transaction Group Max: 0
Data Write Path: ANY
Data Read Path: ANY
Session Initiation: ClientOrServer
High-level Address:
Low-level Address:
Collocation Group Name:
Proxynode Target:
Proxynode Agent:
Node Groups:


***************************************************************
DSM.OPT


NODENAME APPSRV2008
TCPSERVERADDRESS 192.168.1.135




EXCLUDE C:\...\*
EXCLUDE D:\...\*
EXCLUDE E:\...\*
INCLUDE D:\databases\...\*


***************************************************************


I do not have a DSM.SYS in the BACLIENT folder.


***************************************************************


These first 2 nodes are the ones with the wrong contact info:


SQL table mynodes created to map base table Nodes.
SQL table myaddr created to map base table Schedule.Node.Addresses.


Node name - Contact address
------------------ - ------------------
ANDOVER - CONTINUUMDB
APPSRV2008 - INFO_TECH


Where would it be picking up these names from. It should show appserver.gecdsb.local

FYI: There is no activity in dsmsched.log or dsmerror.log from last night, just the night before when it was working properly before it mysteriously changed its address.

 

[ironfang]

So on the clients DSM.OPT file I have added SESSIONINITIATION SERVERONLY
and on the server I did UPDATE NODE APPSRV2008 HLADDRESS=192.168.1.131 LLADDRESS=1501

and this seems to have corrected that entry in that Contact Table.

Will wait for tonight and see if the schedule goes.

 

[ironfang]

OK,
If I Restart the TSM Schedule service on APPSRV2008 the table gets populated with INFO_TECH again!!!

Where is it picking this up from?

 

[ironfang]

I have added TCPCLIENTADDRESS=192.x.x.x to my DSM.OPT file to resolve my problem.