Exchange 2013 Sp1 Loose Truncation
A new feature called “loose truncation” is introduced in Exchange 2013 SP1
When enabling this feature, the database copies will check their own disk space, if disk space becomes low all the databases copies will truncate logs independently. Loose truncation comes into action when the low-space threshold is exceeded (The default is 200GB). In a DAG scenario, where you have a 2 member DAG, out of which 1 node fails, you take a backup of the database but the logs are not moving to the passive node. In such scenarios the log will soon fill up the drive & will dismount the database due to low disk space. Loose truncation concept is very much useful in such scenarios. Where it checks the disk space & truncates the logs automatically. This feature is not enabled by default & it can be enabled by adding Registry keys.
How to configure loose truncation?
As we saw before this feature is enabled by adding registry key. We need to create 3 registry keys. Below is the Registry path:
HKLMSoftwareMicrosoftExchangeServerV15BackupInformation
Keys we need to create:
LooseTruncation_MinCopiesToProtect
LooseTruncation_MinDiskFreeSpaceThresholdInMB
LooseTruncation_MinLogsToProtect
Quick description on what the above keys are:
LooseTruncation_MinCopiesToProtect:
This is used to enable Loose Truncation. We have to specify the number of passive copies to protect from loose truncation on active copy of database. If we have a 2 ode DAG, the value to specify will be 1 (based on my passive nodes). If the value is set to 0 it disables loose truncation feature.
LooseTruncation_MinDiskFreeSpaceThresholdInMB:
This value is about the available disk space in MB that should exist on disk holding transaction logs, before the loose truncation feature is used. Default value is 200MB, which can be modified according to the requirement.
LooseTruncation_MinLogsToProtect:
This key defines the amount of logs that will be protected from truncation. When setting a value to this key, it applies to both the active & passive nodes. On a passive database the number is calculated based on the last available log file.
So how do we know if loose truncation is working?
Well, after enabling loose truncation we should be sure to know if it’s working. To monitor the loose truncation we can look that event viewer for the following events:
Event ID: 170 – When the truncation happens.
Event ID: 848 – for the active database copy
Event ID: 861 – for a passive database copy
Event ID’s: 848 & 864 can help us in detecting serious disk space issue. We can see these 2 events instead of Event 170 when loose truncation is being used.
Ratish Nair
Microsoft MVP | Exchange Server
Team @MSExchangeGuru
