Replication Over a WAN

Directory Server enables you to perform all forms of replication including multimaster replication between machines connected through a wide area network (WAN). This replication allows supplier servers to initialize and update consumers by making optimal use of the bandwidth over networks with higher latency and lower bandwidth.

When deploying or troubleshooting a replication topology that replicates over a WAN, you must check network speed, latency, and packet loss. Network problems in any of these areas might cause replication delay.

In addition, replication data transfer rates will always be less than what the available physical medium allows in terms of bandwidth. If the update volume between replicas cannot physically be made to fit into the available bandwidth, tuning will not prevent your replicas from diverging under heavy update load. Replication delay and update performance are dependent on many factors, including but not limited to: modification rate, entry size, server hardware, error rates, average latency, and average bandwidth.

If you have questions about replication in your environment, contact your Sun Service Provider.

Internal parameters of the replication mechanism are optimized by default for WANs. However, if you experience slow replication due to the factors mentioned previously, you might want to empirically adjust the window size and group size parameters. You might also be able to schedule your replication to avoid peak network times, thus improving your overall network usage. Finally, Directory Server supports the compression of replication data to optimize bandwidth usage.

Configuring Network Parameters

The window and group network parameters determine how the replication mechanism groups entries to send them more efficiently over the network. These parameters affect how suppliers and consumers exchange replication update messages and acknowledgments. The parameters are configurable in every replication agreement, which allows you to tailor the replication performance according to the specific network conditions of each consumer.

Monitor the effects of any modifications that you make and adjust the parameters accordingly. Refer to Getting Replication Status for instructions. You do not need to interrupt replication to modify the window size and group size parameters.

Configuring Window Size

The window size (default value 10) represents the maximum number of update messages that can be sent without immediate acknowledgment from the consumer.

It is more efficient to send many messages in quick succession instead of waiting for an acknowledgment after each message. Using the appropriate window size, you can eliminate the time replicas spend waiting for replication updates or acknowledgments to arrive.

If your consumer replica is lagging behind the supplier, increase the window size to a higher value than the default, such as 100, and check replication performance again before making further adjustments. When the replication update rate is high and the time between updates is therefore small, even replicas connected by a local area network (LAN) can benefit from a higher window size.

To Configure Window Size

You can use DSCC to perform this task. For information, see Directory Service Control Center Interface and the DSCC online help.

1. Modify the window size.

   $ dsconf set-repl-agmt-prop -h host -p port suffix-DN consumer-host:consumer-port transport-window-size:value

   For example:

   $ dsconf set-repl-agmt-prop -h host2 -p 1389 dc=example,dc=com host1:1389 \ transport-window-size:20

Configuring Group Size

The group size (default value 1) represents the maximum number of data modifications that can be bundled into a single update message. If the network connection appears to be impeding replication, increase the group size to a higher value than the default, such as 10, and recheck replication performance.

When increasing the group size, make sure that the following are true:

• The window size is set significantly higher than the group size.

• The window size divided by the group size is much greater than the value for nsslapd-maxThreadsPerConn under cn=config on the consumer (typically twice as large).

  When the group size is set higher than 1, the supplier does not wait to fill a group before sending updates to the consumer.

To Configure Group Size

You can use DSCC to perform this task. For information, see Directory Service Control Center Interface and the DSCC online help.

1. Modify the group size.

   $ dsconf set-repl-agmt-prop -h host -p port suffix-DN \ consumer-host:consumer-port transport-group-size:value

   For example:

   $ dsconf set-repl-agmt-prop -h host2 -p 1389 dc=example,dc=com host1:1389 \ transport-group-size:10

Scheduling Replication Activity

If immediate synchronization between your replicas is not critical, you can schedule replication during periods of low network usage. Replication of data should complete significantly faster when the network is more available.

You can schedule replication to start and end at a certain time of day, on a daily or weekly basis. You can do this independently for every consumer through its replication agreement. The new schedule will take effect immediately, causing the next replication of data for the corresponding consumer to be delayed until first allowed by the schedule.

To Schedule Replication Activity

You can use DSCC to perform this task. For information, see Directory Service Control Center Interface and the DSCC online help.

1. Modify the replication schedule.

   $ dsconf set-repl-agmt-prop -h host -p port suffix-DN \ host:port repl-schedule:value

   For example, if you want to set replication to occur between 2:00 and 4:00 every night, type:

   $ dsconf set-repl-agmt-prop -h host2 -p 1389 dc=example,dc=com host1:1389 \ repl-schedule:"0200-0400 0123456"

   where 0123456 indicate the days of the week, with 0 representing Sunday, 1 representing Monday, and so on.

Configuring Replication Compression

To reduce the bandwidth used by replication, you may configure replication to compress the data that is sent when updating consumers. The replication mechanism uses the Zlib compression library. Both supplier and consumer must be running on a Solaris or Linux platform to enable compression.

You should empirically test and select the compression level that gives you best results for your expected replication usage in your WAN environment. Do not set this parameter in a LAN where there is wide network bandwidth because the compression and decompression computations will slow down replication.

To Configure Replication Compression

You cannot use DSCC to perform this task. Use the command line, as described in this procedure.

1. Configure replication compression on the replication agreement entry in the master server.

   $ dsconf set-repl-agmt-prop -h host -p port suffix-DN \ consumer-host:consumer-port transport-compression:level

   where level can be high, medium, low, or none.

   For example, to use the fastest compression when sending replication updates to the consumer on host1:1389, type:

   $ dsconf set-repl-agmt-prop -h host2 -p 1389 dc=example,dc=com host1:1389 \ transport-compression:high

   For more information about setting the compression level, see Sun Java System Directory Server Enterprise Edition 6.2 Reference.