Semadd Semaphore Error When Starting Database (replication Target)

Hello All,

I am getting below semaphore error when staring database. I have gone through couple of articles for this but unable to resolve this issue.

Code:
09:28:59 BROKER     The startup of this database requires 94Mb of shared memory.  Maximum segment size is 1024Mb.
09:28:59 BROKER  0: Multi-user session begin. (333)
09:28:59 BROKER  0: Connections to this database will not be allowed until all Database Services started have completed their startup and initialization. (10545)
09:28:59 BROKER  0: Before Image Log Initialization at block 0  offset 235. (15321)
09:28:59 BROKER  0: Login by root on /dev/pts/0. (452)
09:28:59 BROKER  0: The OpenEdge Replication Agent is starting...
09:28:59 BROKER  0: Started for bomfglive using TCP IPV4 address 0.0.0.0, pid 23290. (5644)
09:29:05 WDOG       SYSTEM ERROR: Unable to semAdd semaphore set : semid = -32757, errno = 43. (10839)
09:29:05 WDOG       ** This process terminated with exit code 2. (8619)

Kindly give your inputs on this.

Regards,
 

TomBascom

Curmudgeon
The number of semaphores needed is related to your -n startup parameter. If there are multiple databases on the server the total of -n across all databases might also be an issue.

The resolution generally depends on the specific OS that you are running.

It would, of course, also be helpful to know what version of Progress this is -- the .lg file extract looks like a hybrid between v9 and oe10.
 
Hello All, thanks a lot for your replies on this!

Below is the required information:

Code:
Progress Version:

OpenEdge Release 10.2B

OS:

RHEL 6.8 64 Bit


Startup parameters:


$DLC/bin/proserve /path/db-name -n 50 -B 5000 -L 350000 -c 350 -cpstream iso8859-1 -cpinternal iso8859-1 -aibufs 35 -bibufs 25 -bithold 2000 -bistall -ServerType 4GL -spin 50000 -S bomfglive -N TCP -DBService replagent
$DLC/bin/prowdog /path/db-name
$DLC/bin/probiw /path/db-name
$DLC/bin/proapw /path/db-name
$DLC/bin/proapw /path/db-name
$DLC/bin/proaiw /path/db-name

Once I disable sitereplication (target) and removed “-DBService replagent” from startup script then database is starting successfully. One more thing I would like to add here is, Initially I wasn’t able to enable this database as site replication (target) and for solution of that I have copied “rpagent” and “pmmgr.properties” files from source server (took reference from progress articles).

Kindly give your inputs on this.

Regards,
 

Cringer

ProgressTalk.com Moderator
Staff member
It's not likely to be the problem, but a few comments:

1) -B 5000 is very small indeed. How big is your database?
2) -L 350000 do you monitor lock table high water mark? Is this a valid level? If you never get close to this then reduce this parameter. If you do get close to this look at your application.
3) -spin 50000 is far too high. Pi * birthyear of dba is sufficient as a starting point.
 
I totally agree with you @Cringer regarding these parameter. I thought, it might be a memory issue so decreased -B value. Now, I have decreased the value of -L and -spain as well but getting the same issue that i have described above.

Regards,
 
Top