Hi BigFix Community,
I am hoping to get some advice on optimizing our Linux patch deployment times. Currently, our patching baselines for RHEL and Amazon Linux 2023 are taking around 5 to 7 hours to complete. We typically have about 150 components in a single baseline.
Based on our client logs, the delay is caused by BigFix running a separate dnf transaction for every single patch sequentially (taking about 10 minutes per fixlet).
To reduce this time, we are planning to use the native Multiple-Package Baseline Installation "bookend" method to consolidate everything into a single transaction.
Could the experts here kindly review my proposed plan below and let me know if this configuration is correct?
Step 1: Add the Enable Task (Top) I will add the "Enable the Multiple-Package Baseline Installation feature" task as the very first component in the baseline.
- Setting: Uncheck "Baseline will be relevant on applicable computers where this component is relevant".
Step 2: Add the Patches (Middle) I will add our ~150 native Linux patch fixlets directly after the Enable task.
- Setting: Leave their component relevance boxes checked.
Step 3: Add the Execution Task (Bottom) I will add the "Multiple-Package Baseline Installation - Execution" task as the very last component.
- Setting: Uncheck "Baseline will be relevant on applicable computers where this component is relevant".
Step 4: Action Execution Settings When taking action on the baseline, I will apply these two specific settings in the Execution tab:
-
Disable pre-caching: Uncheck "Start client downloads before constraints are satisfied".
-
Force continuation: Check "Run all member actions of action group regardless of errors".
My questions for the community:
-
Are the 4 steps and settings above mapped out correctly?
-
Can this Multiple-Package method successfully handle a large baseline with up to 150 components?
-
Will this effectively bypass the sequential execution bottleneck and reduce our 7-hour deployment time?
Thank you so much in advance for your time and guidance!
