server and server process??

Posted by Aravindhacks on 09-Aug-2013 02:24

i'm confused with the OE server and server process. In sybase,under one server mutiple database can be created.Also multiple servers can be created using sysconfig tool.Like that,can we start multiple OE database server on single host say localhost where OE is installed????

All Replies

Posted by egarcia on 09-Aug-2013 04:16

Hello,

Just a quick reply.

In OpenEdge, each database server (broker) process has one database.

You can create and start multiple databases on a single host.

When you start a database server using the proserve command or from OpenEdge Management/OpenEdge Explorer, the broker process is started.

The broker then start server processes to serve remote/network connections. Local clients can connect directly to the database using shared memory.

Example:

     Client connection using network parameters:

          prowin32 sports2000 -S 20000

          prowin32 sports2000 -H host -S service-name

          prowin32 sports2000 -H host -S port-number -p _dict.p

     Local client connection using shared memory:

          prowin32 sports2000

          prowin32 c:/db/sports2000

          prowin32 sports2000 -p _admin.p

          prowin32 sports2000 -p _dict.p

You can also connect using _progres instead of prowin32.

The database can have multiple servers for server ABL clients and for SQL clients. (Parameters such as -Mn, -Ma can control the number of servers and clients per servers.)

The -S parameter tells the database to accept network connections.

The executable for the broker process and the server is _mprosrv.

The executable for the SQL server is _sqlsrv2.

You can monitor the database using OpenEdge Management via a web interface.

From the command line, you can use the PROMON utility.

The following links to the documentation have information on this:

http://documentation.progress.com/output/OpenEdge112/oe112html/wwhelp/wwhimpl/js/html/wwhelp.htm#href=Data%20Management/Database%20Administration/08dmadmch03.019.3.html

http://documentation.progress.com/output/OpenEdge112/oe112html/wwhelp/wwhimpl/js/html/wwhelp.htm#href=Getting%20Started/Installation%20and%20Configuration/26gsinsapf.27.2.html

http://documentation.progress.com/output/OpenEdge112/oe112html/wwhelp/wwhimpl/js/html/wwhelp.htm#href=Data%20Management/Database%20Administration/29dmadmch21.040.02.html

I hope this helps.

Posted by Aravindhacks on 09-Aug-2013 04:47

Thanks egarcia.

I understood but can we start multiple server instance of OE and able to create database for each instance??

Posted by egarcia on 09-Aug-2013 06:44

Yes, you can start multiple database server instances on a single host.

Notice that it is not that you create a server and then a database for the server instance.

In OpenEdge, you create a database and then start a server for the database.

Perhaps, working from the command line could give you another point of view on how this works.

You can use the Proenv shortcut on Windows or the proenv command on UNIX/Linux to get the environment setup for the command line.

Example:

prodb sports sports

prodb sports2000 sports2000

proserve sports -S 10000

proserve sports2000 -S 20000

The above command will create and start the database broker for two databases.

If you use the following commands on sports2000:

prowin32 sports2000

prowin32 sports2000 -H localhost -S 20000

promon sports2000


The User Control page will show you something like the following:

User Control: by user number for all tenants
  Usr:Ten   Name      Domain     Type      Wait  Table               Dbkey     Trans      PID Sem Srv Login  Time    IPV
# Remote Address
    0       egarcia       0     BROK       --      0                    0         0    11068   0   0 08/09/13 07:21
    1       egarcia       0     SERV       --      0                    0         0     5596   0   0 08/09/13 07:22
    2       egarcia       0     SERV       --      0                    0         0      424   0   0 08/09/13 07:26
    5       egarcia      -3     SELF/ABL   --      0                    0         0    11552   0   0 08/09/13 07:22
    6       egarcia       0     MON        --      0                    0         0     9088   0   0 08/09/13 07:27
   23       egarcia      -3     REMC/ABL   --      0                    0         0        0   0   2 08/09/13 07:26 IPV4 127.0.0.1
   24       egarcia      -3     REMC/ABL   --      0                    0         0        0   0   1 08/09/13 07:22 IPV4 127.0.0.1

In this output, you can see the broker process, two server processes, one self-service client, the promon client, and two remote ABL clients connected to the server processes.

You can configure -Mn,-Ma and -Mi to change the number of server processes that are started per remote clients. You generally do not need to change this. However, I mention it because you seem to be interested on the architecture.

http://documentation.progress.com/output/OpenEdge112/oe112html/wwhelp/wwhimpl/js/html/wwhelp.htm#href=Deployment/Startup%20Command%20and%20Parameter%20Reference/08dpsprch04prm.47.117.html

Another way to get information on the command-line arguments is by using the Procedure Editor (started with prowin32) by entering the argument and highlighting and then pressing F1.


I hope that this clarifies.

Posted by Aravindhacks on 09-Aug-2013 07:13

Thanks buddy.

Yeah it was very helpful.

But Can i able to do replication on the same host where my OE is installed? If so How do i configure my target machine on the same host??? For this is why i need to create two OE server instance,not server process or broker.

Posted by egarcia on 09-Aug-2013 08:18

You are welcome.

I have not played with replication.

My understanding is that you should be able to have both databases on the same machine.

(The source database is accessed directly, and the target database is specified as db, host and port.)

Generally, depending on your goals, the database would be on a different machine.

The OpenEdge Replication User Guide comments on this:

http://documentation.progress.com/output/OpenEdge112/pdfs/ffr/ffr.pdf

"

For added protection against data loss, locate the target database on a different
machine from the source.

"

"Figure 1

also shows the source and target databases as residing on separate
machines, which is the recommended configuration. Setting up the source and target
on separate machines allows the target database to continue if the machine hosting the source database fails.

"

http://knowledgebase.progress.com/articles/Article/P28026?q=replication&l=en_US&fs=Search&pn=1

You could contact Technical Support if you are blocked on setting up this configuration.

I hope this helps.

Posted by Thomas Mercer-Hursh on 09-Aug-2013 08:54

I hope this is only for experimental purposes since having the replicate on the same host for production would be foolish.

Posted by Aravindhacks on 11-Aug-2013 22:17

Yes Thomas. This is only for experimental purpose not on production.Since i am provided with single machine for performing my tasks.

This thread is closed