HP MPX200 Multifunction Router - Manual (Page 141)

Online ETC job
•
If an online (local/remote) job is running while the host is writing to the source LUN, ETC is
the total outstanding blocks that are yet be copied from source to destination plus any
outstanding Dirty Region Log (DRL) blocks, divided by the job’s current performance (MBps):
Outstanding blocks + DRL blocks / Current performance MBps = ETC
Behavior characteristics
•
ETC is calculated every time the job is queried.
•
If current performance is 0, the ETC is displayed as ---
•
If an online job is completed and there are no flush I/Os going on, the ETC is displayed as
---
and current performance is 0.
•
Within the first 30 seconds after a job completes the current performance displays a value,
but the ETC for the job is 0.
•
If an online job is in Copy Complete state, but DRLs are being flushed, both current
performance and values of ETC are displayed.
•
For failed, paused, suspended, and stopped jobs, the ETC is displayed as --- and current
performance is 0.
Best practices
This section provides some best practice recommendations for DMS usage.
When to use offline data migration
For most non mission-critical applications, MPX200 provides effective offline data migration. You
can configure data migration jobs in MPX200 while applications and servers remain online. Each
MPX200 blade can support up to 4 TB per hour (typical midrange storage arrays can sustain
between 2 to 3TB per hour).
Offline migration is more effective than online migration under the following conditions:
•
If the server has less than 1 TB of data to migrate, the overall down time is comparable between
online and offline migration. However, offline migration takes less setup time and provides a
simpler migration process.
•
If you have a large number of small LUNs on the same server, use offline migration.
•
If you have applications with large amount of data (more than 2 TB) and can tolerate
reasonable down time, use offline migration.
High availability and redundant configurations
The following recommendations pertain to HA and redundant configurations:
•
The MPX200 is capable of detecting and performing load balancing over multiple paths. To
achieve the highest data migration rates, zone multiple ports from each array controller with
each port of the MPX200.
•
To maximize performance, zone each MPX200 FC port with one or two ports from one of the
array controllers.
•
To optimize performance, HP recommends that LUNs under migration are balanced across
two controllers of a storage array.
Best practices 141