A replication holding cache is used to store captured or imported events that failed to replicate successfully to a parent server. If a parent server is unable to store a replicated event for any reason, it reports the failure back to the child machine which writes an entry for the ‘failed’ event to the replication holding cache.
By default, the child machine will attempt to resend events in the cache three times. After that, cached events remain in the cache until you manually delete them or reset the cache (that is, restore them to the main replication queue). If required, you can reset the retry limit—see below.
Likewise, as a safeguard against serious or persistent replication failures, the holding cache has a maximum event threshold which, if exceeded, causes the local infrastructure to be suspended. This means, for example, that an import operation is suspended if the CMS is unable to store events arriving from an Event Import server. By default, this threshold is set to 100 events. To reset this threshold, see below.
Note: When you reset the cache on suspended machines, the machines are automatically resumed.
|
Copyright © 2014 CA.
All rights reserved.
|
|