Not seeing preferred number of Tx processes

Using FoundationDB Running FoundationDB
1

We have a 6 node FDB cluster.
(3 i3.xlarge + 3 i3.4xlarge)

3 i3.xlarge(4 cores) have 1 coordinator, 2 transaction and 1 stateless processes defined as preferred.
Snippet from foundationdb.conf file is given below

[fdbserver.4500]
class = coordinator
public_address = :4500
listen_address = public
datadir = /mnt/fdb/4500

[fdbserver.4501]
class = transaction
public_address = :4501
listen_address = public
datadir = /mnt/fdb/4501

[fdbserver.4502]
class = transaction
public_address = :4502
listen_address = public
datadir = /mnt/fdb/4502

[fdbserver.4503]
class = stateless
public_address = :4503
listen_address = public
datadir = /mnt/fdb/4503

All the 3 i3.xlarge instances have above configuration.
3 i3.4xl (16 cores) instances are used as storage nodes. They have 12 SS + 1 coordinator + 3 stateless processes configured on each of them.

As per above conf, I expect 6 transaction process running in the cluster. However we see only 3 transaction processes active at any point. Is this expected behavior?

2

When you say “active”, do you mean that there are only processes serving as logs? If so, that’s controlled through the database configuration. The default is to recruit three logs, even if there are more transaction-class processes serving in the cluster. You can recruit more by running configure logs=6.

3

Thank You @john_brownlee. That helped.

Is there a command that shows current configurations along with the default configurations.
I tried status, but it was not showing default number of logs processes.

4

I’m not aware of any command that shows the defaults. That seems like a worthwhile enhancement, if we can find a way to make it clear and consistent.

5

Thank You @john_brownlee
Yeah, looks like a worthwhile enhancement.