Вы находитесь на странице: 1из 2

target replication server must be accessible from a source replication server by

using an IP connection. The connection must provide sufficient bandwidth to


accommodate the volume of data to be replicated. If the connection is insufficient
and becomes a bottleneck for replication, keeping the data on the two servers
synchronized can be a problem. Keep in mind that you can use client-side data
deduplication with node replication to reduce network bandwidth requirements and
storage requirements

Replicating all nodes, Replicating active data first

Bandwidth es
To calculate the required network bandwidth, determine the following information:
Total amount of data (TD) to be replicated, in gigabytes. To calculate the total
amount of data, see Estimating the total amount of data to be replicated.
Length of replication window time (RWT) in hours. The length of the replication
window time is the time that you schedule during server maintenance for replication
to occur.
Data deduplication ratio (DR), if you are using the data deduplication function.
Issue the QUERY STGPOOL command and use the FORMAT=DETAIL value of to determine the
data deduplication ratio value. If you are not using the data deduplication
function, use 100 as the data deduplication ratio value.
Calculate the bandwidth required by using the following formula:
(TD * (100 / DR) * 8192) / (RWT * 3600) = Required_Network_Bandwidth (Mbps/second)

You can estimate how many hours it will take to replicate data. You must know the
amount of data you want to replicate and the bytes per hour that the network can
process.

Procedure
Determine the time that is required for replication by using the number of bytes
you want to replicate and the bytes per hour value. Use the following formula to
calculate how many hours it takes to complete replication:
Total_bytes_to_be_Replicated / Bytes_Per_Hour = Hours_to_Complete_Replication.

To calculate the Total_bytes_to_be_Replicated and the Bytes_Per_Hour values,


complete the steps in Tuning replication processing.

For the initial replication, determine how many hours it takes for the replication
to occur over the network during a daily window for replication by using the
following calculation:
Days_to_Complete_Replication = Hours_to_Complete_Replication / 24

You can determine how many bytes per hour are replicated over your network at a
specific time by tuning the settings that affect replication and by testing
replication processing. Based on the replication results and the total amount of
node data that you replicate, you can choose which method to use for the initial
replication.

About this task


Do not use a storage pool that is enabled for data deduplication to test
replication. By using storage pools that are not enabled for data deduplication to
test replication processing, you avoid processing extents that can increase the
amount of preprocessing time of the replication process. By determining the data
transfer and network capability of your replication operation without extent
processing, you get a better representation of the capability of your system. Test
replication processing with storage pools that are enabled for data deduplication
if you want to determine the effect of data deduplication on replication
performance alone.

You must calculate the bytes-per-hour value for each source replication server
individually. You can determine which method is the most suitable for the server,
based on its bytes-per-hour value.

Complete the following steps to determine how much data you can replicate during a
specified timeframe so that you can tune replication processing for a server.
Repeat these steps to obtain bytes-per-hour value for each server that you want to
use for replication processing.

Procedure
Complete the following steps to select the appropriate data:
Select one or more nodes and file spaces that have approximately 500 GB to 1 TB of
total data.
Select data that is typical of the data that you replicate on a routine basis.
Select nodes that are configured for replication.
To display the amount of data in a file space, issue the QUERY OCCUPANCY command.
Select a timeframe during which replication is running normally.
If you plan to use Secure Sockets Layer (SSL) as the communication protocol for
replication processing, ensure that SSL is enabled.
Restriction
SSL can slow replication processing.
Start the replication process by issuing the REPLICATE NODE command.
After the replication is completed, review the summary message that is displayed.
Use the Amount transferred and Elapsed time values that display in the summary
message to determine the number of bytes per hour that can be replicated.

Вам также может понравиться