Can someone provide info on from where the data in the srv.dat file (residing on the SD client machine) comes from? Is this data stored somewhere in the SD database which is retrieved and stored in a file called srv.dat when a user logs on to a SD client? i am just guessing.. Can someone provide the exact information?
when a user logs onto OVSD for the first time it stores the server name and any other available servers running at that time in the srv.dat file.
If you add additional servers in the future the users srv.dat file is not updated unless the user actually adds the new server name to th log on screen (basically if they log onto the new server. The file is not dynamically updated by the server to the client.
I'm pretty sure that's not correct Mark. My understanding was that whenever a user connects to HPSD, all the active servers in the REp-Servers table are recorded in srv.dat, and then if you can't connect via one, you will try others.
I checked today at a client - they've only had one app server until about 1 month ago when we added a second. Each of the users I checked has both servers listed in the srrv.dat file, but no-one has modified their user account entries - ie, HPSD has done this for them. Cheers Ken.
Ken I have got from your reply that there is some table in the Service Desk database which holds all data about all the active servers and then whenever a user connects to the SD application server this data is pulled from that table from database to the srv.dat file on the client machine? If this is the case then can you tell me which is that table ()exactly from which this server list is pulled and populated in the srv.dat file?
I stand corrected on this. Note the second paragraph for the condition on this:
Starting Multiple Servers In a multiple server environment, when a client connects to an application server, a list of all application servers that are currently running is sent to the client and stored in the file srv.dat in the user profile. When the application server that the client is connected to fails, the next server on the list takes over.
** However, if, at the moment the client connects to the application server, there is only one server running, and afterwards more servers are started, the client will not know about the other servers. Thus, if the clientsâ server fails, no take-over will take place. This happen because the list is only sent to the client when it first connects at the start of a session. It is therefore recommended to start all your Service Desk application servers in the multiple server environment before clients start to connect. When this happens, the client can be "manually" connected to another application server by changing the hp OpenView service desk Server name field in the account properties dialog box.
Mark - that's interesting. I think that second para must mean within a given session...ie if you have a session open and add more servers then you won't know about them until you stop and restart your client session?
By the way Anu, one of the traps if you restore a Production database to say a Test environment, is that the Rep_Servers table can have the original Production servers in it, and the clients connecting to test can accidentally update Prod data! When we restore a database to dev or test, one of the steps is always to empty rep_severs.
Thank you so much Ken and Mark for your valuable inputs. Yes Mark is right. It is recommended to start all your Service Desk application servers in the multiple server environment before clients start to connect so that when the clients connect its srv.dat file gets populated with data pertaining to all the live SD application server in the environment.
data provided by you has been very useful. thanks once again and Good Night Ken :-)thanks for replying thought it was getting very late.