I'm not sure we could even kill a thread that quickly given the way our tasking framework works. So while we support from an interface perspective the ability to specify a timeout in milliseconds or microseconds, I don't think we could ever enforce that. For example, the following doesn't indicate a timeout took place. But I can't see how it'd be possible that the sync normally completed in 20 milliseconds. $ pulp-admin repo sync --id pulp-f14-64 --timeout milliseconds:20 Sync for repository pulp-f14-64 started Use "repo status" to check on the progress $ pulp-admin repo status --id pulp-f14-64 +------------------------------------------+ Status for pulp-f14-64 +------------------------------------------+ Repository: pulp-f14-64 Number of Packages: 10 Last Sync: 2011-01-13 11:51:53.243000
$ sudo pulp-admin repo sync --id=test1 --timeout=seconds:-22 --help Usage: pulp-admin <options> repo sync <options> Options: -h, --help show this help message and exit --id=ID repository id (required) --timeout=TIMEOUT sync timeout in <units>:<value> format (e.g. hours:2 valid units: seconds, minutes, hours, days, weeks --no-packages skip packages from the sync process --no-errata skip errata from the sync process --no-distribution skip distributions from the sync process -F, --foreground synchronize repository in the foreground As milliseconds and microseconds are removed from valid units in UG and help text for repo sync, this bug no longer exist.