Client is seeing a 20 MB file transfer between primary NMS server Automated Deployment Engine and target machine in less than 1 Minute.
However the ADE if it completes, can take 4+ hours to transfer a 12 MB file.
We are also seeing the following in the Automated Deployment Engine logs:
feb 20 09:53:21:162 INFO [pool-2-thread-2, com.nimsoft.ADE.deployment.deployer.Deployer] Job:1392886400321: Host:abc123.ca.com: Return code for 'mkdir "c:\temp" ' is: 0
feb 20 09:53:23:104 INFO [pool-2-thread-1, com.nimsoft.ADE.deployment.deployer.Deployer] Job:1392886400321: Host:abc123.ca.com: Return code for 'net share robot_share="c:\temp" /GRANT:Administrators,Full' is 0
feb 20 09:53:23:178 INFO [pool-2-thread-2, com.nimsoft.ADE.deployment.deployer.Deployer] Job:1392886400321: Host:abc123.ca.com: Return code for 'net share robot_share="c:\temp" /GRANT:Administrators,Full' is 0
feb 20 09:53:33:117 INFO [pool-2-thread-1, com.nimsoft.ADE.deployment.deployer.Deployer] Job:1392886400321: Host:abc123.ca.com: Copying EXE "../../../install/setup/nimsoft-robot.exe" to abc123.ca.com\robot_share and answer file to bgbulab00.bgbulab.com\robot_share on host abc123.ca.com
feb 20 09:53:33:191 INFO [pool-2-thread-2, com.nimsoft.ADE.deployment.deployer.Deployer] Job:1392886400321: Host:abc123.ca.com: Copying EXE "../../../install/setup/nimsoft-robot.exe" to abc123.ca.com\robot_share and answer file to bgbulab01.bgbulab.com\robot_share on host abc123.ca.com
feb 20 09:58:03:929 FATAL [pool-2-thread-1, com.nimsoft.ADE.deployment.deployer.Deployer] jcifs.smb.SmbException: The handle is invalid.
feb 20 09:58:03:932 INFO [pool-2-thread-1, ADEStatusMessage] Bulletin Update: Status=FAILED, Description=The handle is invalid. jcifs.smb.SmbTransport.checkStatus(SmbTransport.java:563), Host=abc123.ca.com, JobID=1392886400321
feb 20 09:58:03:932 INFO [pool-2-thread-1, com.nimsoft.ADE.deployment.deployer.Deployer] Job:1392886400321: Host:abc123.ca.com: Individual host distribution Ended
feb 20 09:58:34:016 FATAL [pool-2-thread-2, com.nimsoft.ADE.deployment.deployer.Deployer] jcifs.smb.SmbException: The handle is invalid.
Release: UIM 8.x
Component:
There are two files in the attached zip file that need to be placed in the Automated Deployment Engine directory after the Automated Deployment Engine is deactivated.
Once they are in place, reactivate the Automated Deployment Engine and test again.
Speeds should improve.
Summary technical details
We are seeing in the logs that there are timeouts or errors with 'invalid handle' messages which, according to research we have done, basically comes down to certain environments being sensitive to certain factors.
What we have done in this case is provide the jcifis.prp which contains additional options for the JCIFIS system which is the system used to copy the files over. These are the options we've added:
jcifs.smb.client.responseTimeout=30000
jcifs.smb.client.listSize=1400
jcifs.smb.client.listCount=100
jcifs.smb.netbios.soTimeout=10000
It might help to change these numbers - try doubling all of them for example, and see if it makes a difference. Or try cutting them in half and see if THAT makes a difference.
Start with trying the following (add them to jcifis.prp and then restart the ADE probes) and see if this helps:
jcifs.smb.client.maxMpxCount=1
jcifs.smb.client.attrExpirationPeriod=0
jcifs.netbios.retryCount=5
jcifs.util.loglevel=3
Attachments:
Some customers have reported that the above options were not successful, but the below set of options worked:
jcifs.smb.client.connTimeout=60000 jcifs.smb.client.responseTimeout=60000 jcifs.smb.client.soTimeout=4000 jcifs.smb.client.dfs.disabled=true jcifs.smb.netbios.soTimeout=10000 jcifs.netbios.retryTimeout=9000 jcifs.netbios.retryCount=5 jcifs.resolveOrder=DNS