The TrueSight Capacity Optimization (TSCO) Gateway Server nightly Manager runs UDR data transfer failing with the error, "Collected data not successfully transferred from agent computer during the time period allowed" for some machines is usually due to one of two things:
(A) The $BEST1_HOME/local/setup/collectManager.cfg TRANSFER_CONCURRENT_GLOBAL parameter is too low. The default value of 10 is OK for a small environment but is too low for a normal production environment. This value can be safely increased to 25 on pretty much any console hardware and can be increased beyond that if the console has a quality I/O subsystem. Usually, you don't want to get beyond 40 or 50 for this setting. If the value is set too high you can saturate the I/O subsystem during transfer first causing slower transfers then causing significant transfer failure and retry.
(B) Assuming the TRANSFER_CONCURRENT_GLOBAL setting is OK, the TRANSFER_DURATION in the Manager is too low. This can be because there is something causing the computers to transfer really slowly.
One can see that by running a command like this:
$BEST1_HOME/bgs/bin/udrCollectStat -D -d `date --date=yesterday +%m-%d-%Y` -f "%v, %r, %d, %n: %s, %ch, %ce, %ces, %th, %te, %tes, %tg, %tt"
Look at the last column in the output for high "Transfer Time" (%tt). So, if some computers being transferred are taking a really long time (say an hour) that computer will own a transfer slot for that whole hour which could back up other computers waiting to transfer. You don't see that very commonly except when there is a Manager run on the console across a very slow WAN link.
Increasing one or both of those parameters will pretty much always resolve that error.
