Does this happen to any machine you push to? Could the client PC be in Gateway mode?
Sounds like machines are in Gateway Mode See
Run C:\Program Files\LANDesk\LDClient\BrokerConfig.exe
Gateway Information Tab
Make sure Connect Using the Management Gateway is Not checked.
If these client are mobile and are configured for the gateway they may be having issues contacting the core and think they are external.
Yeah...this is common when an agent workstation cannot resolve the Core
name or connect to it. The agent workstations thinks they need to use
the gateway because the Core is inaccessible.
Nope... The machine is not set to gateway mode and it can see the server.
I have re-installed the client and tried again with the same error.
What if you run brokerconfig and tell it to always you the core? Also, in your client config how do you have your core server specified (netbios name, FQN, IP, etc...)?
No change. I switched it to always use the core and got the same error.
The part that is really bugging me is that the old server name is still in there.. I can't find any ref to it anywhere in my LanDesk config.
Check your distribution package to make sure it is still not referencing your server - rdgto09. Reset the hash. If it is a script/batch make sure it is not hard coded there. Also check you old preferred server names and remove the old server.
Where is the preferred server name?
It's the line 'http://LANDesk.Gateway@10.1.2.3/ldlogon/FileLists/taskmanifest.SMSSERVER.94.18.ini' that references the old IP and old servername. Very odd.. The package was built after the move!
Configure -> Prefered server.. Sorry.. Yes I updated that a long time ago..
Ok, open up the METASYSTEMS table. What are the SYSTEMNAME and CORESERVER?
And just for S&Gs on the core HKLM\software\intel\landesk\ldwm\Coreserver
Also, just to take remote consoles out of the picture - create a new task using the console on the core and see if you get the same thing.
If they all point to the correct server, then I would open up a case.
LDMS 8.7 SP3 | 3 Cores - 2003| 16,000 nodes | SQL 2000 SP4 | Suggested Posting Guidelines
I am not an employee of LANDesk. My views expressed in this post are my own and do not necessarily reflect the views of any normal human being.
SYSTEMNAME was wrong.. Changed it, built a new package.. But still the old server name sits there
How to troubleshooting the Taskmanifest file failing to download?
Failed to download all additional files and folders
Wed, 29 Aug 2007 02:12:50 .\AdditionalFiles.cpp(39): (8DAC4026): Failed
to download file
Wed, 29 Aug 2007 02:12:50 Client thread failed to start
1. Make sure the Core Server is on the latest service pack.
2. From the agent workstation, try to download the Taskmanifest file
as local system manually as directed below:
and sdclient.exe /f
3. Get the sdclient_task##.log
4. Xtrace lddwnld and sdclient.exe
Get these Xtrace logs.
5. Get the proxyhost.log
6. Make sure that the proxyhost.log says that the result was 200 for
the web request.
2007-01-19 15:55:47(1692-952) proxyhost.exe:Made direct
(non-proxy) connection to STL-NT-LD1:80
2007-01-19 15:55:47(1692-952) proxyhost.exe:127.0.0.1:3022
Connection close 0 0 0 0
2007-01-19 15:55:47(1692-952) proxyhost.exe:127.0.0.1:3022 -
LANDesk.Gateway [19/Jan/2007:09:55:47 -0600] "HEAD
98.ini HTTP/1.1" 200 97 354
2007-01-19 15:55:47(1692-952) proxyhost.exe:127.0.0.1:3022 EOS on
7. Delete the taskmanifest for the task. Reset the Package Hash,
restart the scheduled task.
8. Find out what IP Addresses the Core server has. Does it have only
one, or does it have multiple IP Addresses? Can the client hit the Core
Server on all IP Addresses.
9. Look in the default agent configuration. Under Standard LANDesk
agent, what is configured for Core Server. NetBIOS name, FQDN, or IP
Address. Can the client hit the Core Server by that value? Does it work
if you change this to a different value (if it is NetBIOS change to
The taskfiles are there and accessible on the correct LanDesk server. (Interestingly the files are called SMSServer).
It's not that the files are not there.. It's that it is looking in the wrong place.. My CORE Server is not 10.1.2.3 and it is not called SMSSERVER.
Also.. The MultiCast looks to be working! I can see the files in the LanDesk folder on the client! Also can see the MultiCast running on the server!
Was SYSTEMNAME the name of the old server? If so I'm wondering if the old name is embedded anywhere else - client configs, etc.... Also, I would restart all LANDesk services, and run coredbutil with build components. I would also be leery about the install/migration with the SYSTEMNAME being wrong. After all that I would recreate the distro package, sched task, and check your client config for the old name. After you recreate the distro package and sched task, etc... look in
<core>\ldmain\ldlogon\FileLists for the approriate name and hopefully it will have the correct server name taskmanifest.<Correct Core Name>.xxx.xxx.ini