I would GREATLY suggest that you attempt to find a way to just run the
required sql against the database directly in the first recordset and not
have to build your own query at all. I know in a lot of companies, pushing
through sql changes can sometimes be a really big pain, but I think it would
remarkably improve your performance. The query functions just aren't made to
handle the number of rows that you're talking about, and you're basically
taking something from a nice cursor driven data stream coming from the
database and slamming it directly into ram, with a lot of CPU overhead, and
slowing down the read time of the large data stream from the database, and
it just might be the database's connection timeout settings that you're
running into while you're attempting to process the data.
 

Bruce Dunwiddie 
Ticket Technology 
P: 866.543.3331 
F: 913.451.7832 
[EMAIL PROTECTED] 

-----Original Message-----
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]
Behalf Of Benjamin Price
Sent: Thursday, October 09, 2003 10:47 AM
To: [EMAIL PROTECTED]
Subject: RE: [KCFusion] JRun issue when filling Verity on MX


I create another Query using the QueryNew, QueryAddRow and QuerySetCell
functions.  As I loop over the first recordset, I compact several rows at a
time into one new row.
 
 
-----Original Message-----
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On
Behalf Of Bruce Dunwiddie
Sent: Thursday, October 09, 2003 12:29 PM
To: [EMAIL PROTECTED]
Subject: RE: [KCFusion] JRun issue when filling Verity on MX



What do you mean exactly by saying that you're creating another query object
from the first recordset? Are you using query of query to create another
recordset, or are you looping over the first recordset to dynamically
generate the sql for the second query?
 

Bruce Dunwiddie 
Ticket Technology 
P: 866.543.3331 
F: 913.451.7832 
[EMAIL PROTECTED] 

-----Original Message-----
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]
Behalf Of Benjamin Price
Sent: Thursday, October 09, 2003 10:32 AM
To: [EMAIL PROTECTED]
Subject: [KCFusion] JRun issue when filling Verity on MX


Interesting issue involving several ColdFusion topics here.
 
We have upgraded to MX 6.1 and have been having numerous issues. The current
issue happens when I try to refresh a Verity collection. The code to fill
the collection takes quite a while to complete because I have to loop over
40,000+ rows of data. I create another query object from that recordset
which ends up about 11,500 rows. I then put that into the Verity collection.
The problem is that the page which does all this is returning an error which
gives no clues as to where in the code the issue happens. The error is this:

Server Error
The server encountered an internal error and was unable to complete your
request. 
JRun closed connection.

The jrun log isn't helpful either as this is the only entry when the above
error happens:

2003-10-09 07:05:57 jrISAPI[2816:1228] command error: Connection reset by
peer <- [8192]
2003-10-09 07:05:57 jrISAPI[2816:1228] returning error page for Connection
reset by peer

Prior to the timing out and error listing, the jrun.exe process has been
running from 50 to 100% of the processor, yet it never even get close to the
alotted memory (set in CF Admin to 1512K), maxing out at about 512K. 

We are running this on a Dell server(PowerEdge 6600), dual 1.6 Gig Xeon
processors, Win2K server, IIS 5.0, CFMX 6.1.

One other question we did have is that the build listed for the JRun server
is 63961, from July. The last we can find on this site is from June
(something just over 62000). Could this be the problem?

Any ideas here would help.

 
 

Find out why more Distributors are using DistributorCentral than ASI or
SAGE!

Ben Price

Web Developer

DistributorCentral

913-856-4130

[EMAIL PROTECTED]

 

<<attachment: winmail.dat>>

Reply via email to