Handle potential log wrapping

The potential for log wrap occurs when Enterprise Replication log processing lags behind the current log and the Enterprise Replication replay position is in danger of being overrun.

There are two log positions you should be aware of: the snoopy log position, which is the log position that keeps track of transactions being captured for replication, and the log replay position, which is the log position that keeps track of which transactions have been applied.

A potential log wrap situation is usually caused by the logical logs being misconfigured for the current transaction activity or by the Enterprise Replication system having to spool more than usual. More-than-usual spooling could be caused by one of the following situations:

A one-time job might be larger than normal and thus require more log space.
One of the target servers is currently unavailable and more spooling of replicated transactions is required.
The spool file or paging space could be full and needs to be expanded.

You can configure how Enterprise Replication responds to a potential log wrap situation by specifying one or more of the following solutions, in order of priority, with the CDR_LOG_LAG_ACTION configuration parameter:

Block user transactions until Enterprise Replication log processing advances far enough that the danger of log wrapping is diminished. Blocking user transactions prevents the current log position from advancing. This solution increases user response time. When user transactions are blocked, event alarm 30 unique ID 30002 is raised and the following message appears in the online log:
```
DDR Log Snooping - DDRBLOCK phase started, userthreads blocked 
```
Compress the logical logs and save them to a log staging directory. Log files in the staging directory are deleted after they are no longer required by Enterprise Replication. You must specify the location and maximum size of the log staging directory. This solution uses very little additional disk space to temporarily save log files until the danger of log wrapping is over. The staged log files are deleted after advancing the log replay position.
If log staging is configured, Enterprise Replication monitors the log lag state and stages log files even when Enterprise Replication is inactive.
Dynamically add logical logs. This solution requires enough free space to be available in the logical log dbspace to add dynamic logs. You can specify how many dynamic logical logs to add. You must manually drop the dynamic log files when the danger for log wrapping is over.
Ignore the potential for log wrap. This solution shuts down Enterprise Replication when an overrun of the snoopy log replay position is detected. Enterprise Replication continues to function if the log replay position is overrun. If the snoopy replay position is overrun, Enterprise Replication is stopped, event alarm 47 is raised, and the following message appears in the message log file:
```
WARNING:  The replay position was overrun, data may not be replicated.
```
If the replay position is overrun, restart Enterprise Replication with the cdr cleanstart command to reset replay position to current log position and synchronize the data.
Stop Enterprise Replication on the affected server as soon as it is detected that the log replay position is running behind. When you are ready to restart Enterprise Replication it is necessary to run the cdr cleanstart command only if the log replay position was overrun.

For example, you can specify that during a potential log wrap situation, Enterprise Replication stages compressed logical logs. If the log staging directory reaches its maximum size, then logical logs are added. If the maximum number of logical logs are added, then Enterprise Replication blocks user transactions. Not all options can be combined together in every possible priority order. For example, specifying to stop Enterprise Replication, to ignore the potential for log wrap, or to block user actions must always be either the only option or the last option in the list.