Nope. I'm only using cachedWithin.
This is the first morning I've come in to find the app still running
overnight. There was another string of events in the logs early this AM,
but apparently not enough to swamp the app or Tomcat this time. I can
probably attribute this to either reduced overall demand. This site gets
a critical volume of admission-related transactions leading up to a
monthly event.
/WEB-INF/bluedragon/work/cfquerybatch/ shows that my system is cranking
out 150Mb of "java.lang.NullPointerException" lines per day.
If my queries are not actually being cached, THAT would explain why the
traffic spike would kill this app from the database up.
There's also this in bluedragon.log. But this error has been well
documented in prior threads by myself and a few others. It's the
infamous mystery datasource that exists only by exception (ie there is
no blank entry in the xml), so I don't think it applies:
18/08/11 09:11.32: -] PersistDataJDBC.configure: [--Catch Data--]
type: Database
tagname:
template:
Line: ; Column:
ErrorCode: errorCode.sqlError
message: General SQL Error
detail: The datasource could not be found or was invalid
Al
On 8/18/2011 11:41 PM, Alan Williamson wrote:
Oh this is good information.
Are you using background=true in your query?
--
http://alan.blog-city.com/
-----Original Message-----
From: Alan Holden<[email protected]>
Sender: [email protected]
Date: Thu, 18 Aug 2011 19:00:37
To:<[email protected]>
Reply-To: [email protected]
Subject: Re: [OpenBD] Re: SQL Server Connection Issue only in v1.4
One more chunk of info to add to this mystery:
This OpenBD app has a 10Mb log file called querybatch.log - which is
filled with 207,159 repeating lines: java.lang.NullPointerException
Are OpenBD-specific cfquery tag attributes (like cachename) required?
Al
On 8/18/2011 9:45 AM, Alan Holden wrote:
The error continues to occur, about twice a day or as volume goes up.
The app appears to choke on unfulfilled requests - or open connections
or something: "When connecting to the Database this error was
reported: failed to verify remote server/socket".
The Postgres 7.3.4 database is still working, as the CFMX 6.1 server
continues to use it without issue. All the machines involved are on
the same rack, there's only a router between them. The datasource is
set Per-Request Connections=false, Max Connections=24, Timeouts=20.
It looks like the Tomcat apps get hosed as well... the OpenBD
administrator for THIS deployment stops responding, and OpenBD
administrators for OTHER deployments start throwing directory and file
not found errors... weird stuff.
Would you suggest I try a newer Postgres driver? OpenBD 1.4 is on
postgresql-8.4-701.jdbc4.jar - and the latest is
postgresql-9.0-801.jdbc4.jar. Besides dropping the newer jar into
/WEB-INF/lib - what else needs to be changed?
Was there any general improvements to OpenBD 1.5 in this area? Would
you suggest trying that jar as the next step?
--
official tag/function reference: http://openbd.org/manual/
mailing list - http://groups.google.com/group/openbd?hl=en
--
official tag/function reference: http://openbd.org/manual/
mailing list - http://groups.google.com/group/openbd?hl=en
