Replica State & Ring Consistency Checks
The replica state check lists partitions and the states of the replicas stored in the current Directory Information Base (DIB). Although not likely to happen it is a good idea to periodically check to ensure that replica states have not suddenly changed from ON to some other undesirable state, such as Dying.
You should also occasionally check that each partition replica ring is consistent, that is, the replica ring list on every server holding the same partition should be the same.
To check for ring consistency you run ndsrepair on the server with the master replica, check the ring, then run ndsrepair, on other servers in the ring and check to see if the list and type of replica reported is the same.
For Solaris or Linux execute ndsrepair –P – login as root required – Select Option 6 to check replica state, Select Option 10 to view the replica ring information.
Schema Synchronisation and Status Check
All servers in the eDirectory tree have a copy of the schema, even if the server does not hold any replicas, the schema information must be identical.
When schema changes are made they are applied to the server holding the Master replica of root, from there the changes are propagated down the tree from one server to another. The eDirectory schema synchronisation process is event driven and starts 10 seconds after a schema change is made.
To use ndstrace to check whether the schemas in a tree are in synchronisation:
For Solaris or Linux execute the following:
set ndstrace = on
set ndstrace = nodebug
set ndstrace = +scma (Enables schema sync messages)
set ndstrace = +scmd (Enables detailed schema sync messages)
set ndstrace = *ss (Forces Schema Sync to run)
You can view the trace information in /var/nds/ndstrace.log file, which can be viewed through a text editor. You are looking for “All processed = YES” statement. Or you can use the grep command to interrogate the file for the statement you are looking for.
The replica state check lists partitions and the states of the replicas stored in the current Directory Information Base (DIB). Although not likely to happen it is a good idea to periodically check to ensure that replica states have not suddenly changed from ON to some other undesirable state, such as Dying.
You should also occasionally check that each partition replica ring is consistent, that is, the replica ring list on every server holding the same partition should be the same.
To check for ring consistency you run ndsrepair on the server with the master replica, check the ring, then run ndsrepair, on other servers in the ring and check to see if the list and type of replica reported is the same.
For Solaris or Linux execute ndsrepair –P – login as root required – Select Option 6 to check replica state, Select Option 10 to view the replica ring information.
Schema Synchronisation and Status Check
All servers in the eDirectory tree have a copy of the schema, even if the server does not hold any replicas, the schema information must be identical.
When schema changes are made they are applied to the server holding the Master replica of root, from there the changes are propagated down the tree from one server to another. The eDirectory schema synchronisation process is event driven and starts 10 seconds after a schema change is made.
To use ndstrace to check whether the schemas in a tree are in synchronisation:
For Solaris or Linux execute the following:
set ndstrace = on
set ndstrace = nodebug
set ndstrace = +scma (Enables schema sync messages)
set ndstrace = +scmd (Enables detailed schema sync messages)
set ndstrace = *ss (Forces Schema Sync to run)
You can view the trace information in /var/nds/ndstrace.log file, which can be viewed through a text editor. You are looking for “All processed = YES” statement. Or you can use the grep command to interrogate the file for the statement you are looking for.
No comments:
Post a Comment