2012-02-07 08:07 AM
Is it possible to setup a protection Manager job that can SnapMirror data from one controller to another controller and then to a third controller? I know you can do a snapmirror from primary to seconary and then do a snapvault backup from secondary to terciary but I am wanting that snapvault backup to just be a snapmirror for additional DR.
The basic setup we have is a controller in Europe which does a snapmirror of its data to a controller here in US. I then want to mirror the data on the US controller to another controller here in he US. I know there is also a snapmirror job that mirrors to two destinations but because of the physical location of the controllers, we do not have the bandwidth to mirror all the data multiple times from overseas.
Solved! SEE THE SOLUTION
2012-02-07 08:15 AM
I hope using the protection policy “Chain of two mirrors” should satisfy your requirement.
# dfpm policy list "Chain of two mirrors"
Id Name Policy Type Description
---- ------------------------------------------------------- --------------- ----------------------------
73 Chain of two mirrors Data Protection Local backups on primary, then mirror primary to secondary, then mirror secondary to tertiary
2012-02-22 07:26 AM
Ok, next question.. I am using the Chain of Two mirrors Protection Policy and that works great. Only issue I now have is I'm getting a Lag error on the local backups. Since this is just a snapmirror job, I'm not sure why it cares about local backup lag, but it does. I've tried setting the Error threshold to a shorter time than the warning threshold, which is what it says to do to suppress the error messages, but this does not work.
Any ideas on how to get rid of these Lag errors on the local backups?
2012-02-22 07:32 AM
PM is looking for a snap it created on the primary volume and doesn't find one, hence the error. I've had it a number of times before, and what I usually do to get rid of it is to add in a local snapshot schedule just to take one snap per day, and keep it for a day. This means PM has its own snap to use to monitor the source and should get rid of this error.
2012-02-22 07:51 AM
As Chris said, either add a local backup schedule to your policy or disabling the lag for the root node by changing the following setting in the policy.
Note:This can only be done using cli.
[root@ ~]# dfpm policy node set
Valid options are
nodeName Node Name
hourlyRetentionCount Hourly Retention Count
hourlyRetentionDuration Hourly Retention Duration
dailyRetentionCount Daily Retention Count
dailyRetentionDuration Daily Retention Duration
weeklyRetentionCount Weekly Retention Count
weeklyRetentionDuration Weekly Retention Duration
monthlyRetentionCount Monthly Retention Count
monthlyRetentionDuration Monthly Retention Duration
backupScriptPath Backup Script Path
backupScriptRunAs Backup Script Run As
failoverScriptPath Failover Script Path
failoverScriptRunAs Failover Script Run As
snapshotScheduleId Snapshot Schedule Id
snapshotScheduleName Snapshot Schedule Name
lagWarningEnabled Warning Lag Enabled
lagWarningThreshold Warning Lag Threshold
lagErrorEnabled Error Lag Enabled
lagErrorThreshold Error Lag Threshold
2012-02-22 08:13 AM
C:\Documents and Settings\applied>dfpm policy node get 4598
Node Id: 1
Node Name: controller1
Hourly Retention Count: 2
Hourly Retention Duration: 0
Daily Retention Count: 2
Daily Retention Duration: 0
Weekly Retention Count: 1
Weekly Retention Duration: 0
Monthly Retention Count: 0
Monthly Retention Duration: 0
Backup Script Path:
Backup Script Run As:
Failover Script Path:
Failover Script Run As:
Snapshot Schedule Id: 0
Snapshot Schedule Name:
Warning Lag Enabled: Yes
Warning Lag Threshold: 1209600
Error Lag Enabled: Yes
Error Lag Threshold: 604800
Node Id: 2
Node Name: controller2
Node Id: 3
Node Name: controller3
2012-02-22 09:03 AM
As you dont have any schedule attached to the Primary node disabled the below as mentioned by Adai.
dfpm policy node set <Policy Name or Id> 1 lagWarningEnabled=No
dfpm policy node set <Policy Name or Id> 1 lagErrorEnabled=No