server or the system has no more resources

[rajesh4you]

Hi All,

I am getting the below error when trying to connect a DB. Please anyone throw some light.

"The server or the system has no more resources. Please contact Progress Technical Support. (748)"

 

[Cringer]

I'd suggest reading up on the error on the knowledgebase.

 

[TomBascom]

It is always helpful to reveal the version of Progress and the platform involved.

Usually this would mean that you have exceeded the maximum number of allowed connections or that one of the connection parameters is too small. If the number of connections being allowed is limited on purpose (perhaps because of a license requirement or a design constraint) then you are simply out of luck. If it is misconfiguration of connect parameters then you should review the -n, -Ma and -Mn server startup parameters.

If the database is on a UNIX server then it is also possible that you have exceeded a kernel limit such as number of semaphores or swap space.

 

[bohrmann]

Hi,

I have the same problem, but none of the alternatives listed above explains to me why we encountered this issue. As far as I know neither Kernel parameters nor startup parameters have been changed recently, despite from day to day broker became inactive. I can see the following logs:

S-0001>(23-May-11 15:06:20:451) [5797] Connected to database <db1>, user number 34. (9543)
S-0001>(23-May-11 15:06:20:497) [5797] Connected to database <db2>, user number 13. (9543)
S-0001>(23-May-11 15:06:20:498) [5797] WTA server initialising. (8835)

S-0002>(23-May-11 15:06:21:435) [5798] Connected to database <db1>, user number 33. (9543)
S-0002>(23-May-11 15:06:21:514) [5798] Connected to database <db2>, user number 12. (9543)
S-0002>(23-May-11 15:06:21:515) [5798] WTA server initialising. (8835)

S-0003>(23-May-11 15:06:22:424) [5801] Connected to database <db1>, user number 32. (9543)
S-0003>(23-May-11 15:06:22:429) [5801] WTA server initialising. (8835)
The server or the system has no more resources. Please contact Progress Technical Support. (748)

S-0003>(23-May-11 15:06:22:430) [5801] Disconnecting from database <db1>, user number 32. (9545)
L-4100>(23-May-11 15:06:22:445) Log Closed

It looks like the thread no. 3 is not able to connect to <db2> which is found on another server. This remote server is operating well, also the service has been defined properly in startup file for a long time (also working fine so far). But I'm not sure that I'm on the right clue.
The OS is Sun Solaris, Progress version is 9.1E.

Any idea what could be done?

Thanks,
Peter

 

[RealHeavyDude]

Do you see any messages in the log file of database db2 when the agent #3 tries to connect to it?

Not having changed any configuration settings in the environment does not necessarily mean that nothing has changed in the environment. For example the number of users connecting the database has increased over time and now you are hitting a limit because some of your configuration settings are not appropriate any more.

Heavy Regards, RealHeavyDude.

 

[bohrmann]

Thanks, great idea, the db log on the remote server proves that the problem was with accessing db2.

15:06:20 SRV 4: Login usernum 13, userid dbapro, on <server103> batch. (742)
15:06:20 SRV 4: Received RECONNECT from WTB
15:06:20 SRV 4: Previous message sent on behalf of user 13. (5512)
15:06:21 SRV 4: Login usernum 12, userid dbapro, on <server103> batch. (742)
15:06:21 SRV 4: Received RECONNECT from WTB
15:06:21 SRV 4: Previous message sent on behalf of user 12. (5512)
15:06:22 BROKER 0: No servers are available. Try again later. (1154)
15:08:22 SRV 4: Usernum 12 terminated abnormally. (794)
15:08:22 SRV 4: Logout usernum 12, userid , on <server103> batch. (739)
15:08:22 SRV 4: Usernum 13 terminated abnormally. (794)

Based on the help there could be too many users connecting to the db, but it sounds strange, because it's just a test server, with quite less traffic. Or can be anything else behind this message (1154)?

Regards,
Peter

 

[tamhas]

Just because there aren't a lot of people using it, doesn't mean there can't be more connections than are set for -n .... or, more appropriately, that the number of servers and user count per server aren't set to create a limit lower than -n. Why don't you tell us the parameters with which that DB is being started.

 

[bohrmann]

The -n parameter is set to 9. But it's been the same for several years, and we haven't faced this issue so far.
Anyhow, I restarted the database and it resolved the problem.
Thanks to all of you for your help.

Regards,
Peter

 

[RealHeavyDude]

The -n start parameter defines the number of processes that are allowed to connect to the database - which includes all kinds of clients and background writers, secondary login brokers and so on.

A value of 9 seems odd to me as this means that you deliberately set it lower than the multi-user default - which is 20. Plus 9 seems to me that it is not just some guess - I suppose it is calculated and maybe there is no safety included. By safety I mean that there are things which can happen that make the database think that a client that has disappeared is still connected and therefore is blocking a legitimate connection request from another client as -n is exhausted. Shutting down the database will ultimately solve the problem - but shutting down the database whenever this issue pops up may not be an option that is available all the time, especially in production environments.

Given that less information you provide it's subject to speculation as to what remediation steps you could take besides increasing the -n parameter.


Heavy Regards, RealHeavyDude.