Schedule a transport / enter production instance manually

[Sm1tje]

Hi all,

like in most cases our production environment consists of several application servers and a central instance. Now, when we want to schedule a transport into production the following happens.

Upon scheduling a job is created in the production environment, so far so good. However, when we leave it like this and don't take any additional action the actual import fails with a general error message: an internal error occurred.

To avoid this from happening, after scheduling the transport from Development, we have to logon to the production environment and manually change the job and enter the central instance (server name) as Exec. target.

Is there a way or some setting we have to make / change in order to not having to logon and manually change the server name?

Thanks in advance.

Kind regards,

Micky.

[yakcinar · Accepted solution]

Hello Micky,

Is your prodcution system is HA Installation or Distributed system?

Regards,

Yuksel AKCINAR

[Sm1tje · Accepted solution]

Hi Yuksel,

I'm not a basis consultant myself I'm just the one who ran into this issue after scheduling a transport into production. But I made some requirements and they assured me that we do not have a HA installation. So it must be a distributed system.

Regards,

Micky.

[yakcinar · Accepted solution]

Hello Micky,

There could be error about DIR_TRANS.

I think the directory (..\usr\sa\trans) is on central instance.

If configuration is not done accordingly App servers donot check the global trans dir. They control their local transdir.

Ask BASIS guys to configure DIR_TRANS that can be reached from all app servers. They must see the same address. Could be like \\<Central Host\trans.

Regards,

Yuksel AKCINAR

[Sm1tje · Accepted solution]

Hi Yuksel,

based on your answer I will ask our basis consultant. But I did already check in AL11 (/usr/sap/trans). I can only see the data files (transport files like D<sysid>K9xxxxx in the directory of the central instance. The same directory is also available on the (other) application server but there this directory is empty.

Should the files be transported to both directories when importing them into production? If yes, this would explain this behavior. Or should the TP protocol always check the directory of the central instance? Is that the way to configure this normally?

Kind regards,

Micky.

[yakcinar · Accepted solution]

Hello Micky,

That's what I mean.

Basis guys must configure all instances to see the same trans folder.

On all instanses the user running the sap service SAPSerciceSID must have the same password.

A domain user is recommended for this kind of distributed syatems.

Regards,

Yuksel AKCINAR

[Sm1tje · Accepted solution]

Yuksel,

I had one of basis administrators have a look at it, but they were not able to figure this one out yet. I will share your suggestion with them and see what happens.

Micky.

[Former Member]

Can you share more details about internal error occured (full log with error message as minimum)?

[Sm1tje]

Hi Roman,

no I don't. The last error we had was way back and no more logging is available. As soon as someone of us forgets to change the job manually I will add the error message.

Kind regards,

Micky.