Unit for teiid buffer

[marc.kusters]

What is the unit for the following configuration item?

max-buffer-space

I assumed it would GB, but i'm not sure since documentation is lacking.

Reason I need this to be a higher value is I get the following error described on this page:

Max estimated size exceeded in JDV - Red Hat Customer Portal

More info I found was here:

Memory Management - Teiid 9.0 (draft) - Project Documentation Editor

The thrown error is frustrating because I have got enough disk space and memory available and it is almost not being consumed while the query is running. The query stops at about 6 minutes in. I think this is the first time it happened after upgrading from JDV 6.3.8 to 6.4.5.

TEIID30019 Unexpected exception for request F/7eiYhtWvEM.40: org.teiid.core.TeiidComponentException: TEIID31261 Max estimated size 12,568,686,806 for a single operation/table id 21,736 has been exceeded. The server may need to increase the amount of disk or memory available, or decrease the number of max active plans.

I now solved the problem by just adding an extra zero behind the configuration item .

Another question:

After the above mentioned upgrade a certain query won't run. It complains about strings that contain certain characters; specifically those of some foreign cities that contain e.g. an umlaut or other characters. Is there anyway to fix this?

[shawkins]

> I assumed it would GB, but i'm not sure since documentation is lacking.

The upstream docs are at: Memory Management · GitBook , but yes it appears that it does not directly mention the units.  It is in MB.  

> The thrown error is frustrating because I have got enough disk space and memory available and it is almost not being consumed while the query is running.

See [TEIID-5525] add a flag to revert to the prior behavior (TEIID-4557) - JBoss Issue Tracker - there was an error in the calculation of the estimated size in that it did not take into account serialized / vs heap storage.  That has been corrected and for 6.4.6 the enforcement will be turned off by default.

> After the above mentioned upgrade a certain query won't run. It complains about strings that contain certain characters; specifically those of some foreign cities that contain e.g. an umlaut or other characters. Is there anyway to fix this?

That will take some more specifics to determine what is occurring.  It may be related to [TEIID-5313] Oracle translator issue with mixing string types and general issues with non-ascii strings - JBoss Issue Tr…  which had a follow on issue of [TEIID-5377] Type mismatch when issuing recursive common table expression query - JBoss Issue Tracker

[marc.kusters]

Thanks for the reply!

To what value would you suggest I set:

org.teiid.maxSessionBufferSizeEstimate

[shawkins]

org.teiid.maxSessionBufferSizeEstimate is in bytes and defaults to not enforced.  The value unfortunately is based upon the heap only estimate, so if you do set it you have to base it upon some combination of heap/ fixed buffer and disk - which you need to scale somewhat see System Properties · GitBook.