Ok for all of those following this thread, we resolved our hanging instance
problem!
It was an Efflare imageCR3 service (an image tag alternative to cfimage), that
was getting hung. As a result, it caused any pages that called the tag to hang.
Restarting that service solved the problem.
I have a problem with a non-CF-related out-of-control service on one of
my servers. I purchased a product called ProcessLasso (cost about $150
IIRC). It allows you to assign thresholds for CPU utilization that will
automatically kill rogue processes or restart them. Might be worth
in case ur using windows 2008 server, you have this functionality
built right in :-)
On Sat, Sep 24, 2011 at 1:08 AM, Carl Von Stetten
vonner.li...@vonner.net wrote:
I have a problem with a non-CF-related out-of-control service on one of
my servers. I purchased a product called ProcessLasso
Russ,
I didn't know that. Nice!
Carl
On 9/23/2011 5:11 PM, Russ Michaels wrote:
in case ur using windows 2008 server, you have this functionality
built right in :-)
On Sat, Sep 24, 2011 at 1:08 AM, Carl Von Stetten
vonner.li...@vonner.net wrote:
I have a problem with a non-CF-related
have you checked the currently running requests and performed stack
traces to see what is happening when the instances are hung ?
one common cause for such behaviour is connections to external
resources failing or timing out such as databases.
--
Russ Michaels
www.bluethunderinternet.com :
I have run into similar situation long time back and I have 2 solutions for
you. In my case this was due to memory leakage. The memory was used up
completely in 2-3 days and required server restarts.
1. Use Java Garbage Collection
Modify the JVM.CONFIG file and and append the following line to
Yes, we use Fusion Reactor and can view the vitals including which threads are
hung (essentially all threads). Other vitals don't point to anything suspect.
However today, we had a different problem. Only one instance hung and it was a
CPU problem. The entire server memory was being used by
The BIG question here is that while the CPU(s) were maxed - was there a long
running thread at the same time? If so, look into what the thread was doing.
If not, then figure out what else on the server is making heavy use of the
CPU(s).
Wil Genovese
Sr. Web Application Developer/
Systems
Right, couldn't even get there as the mouse was not responding since the cpu
was at 100%. Was only barely able to restart the service.
~|
Order the Adobe Coldfusion Anthology now!
Thanks for the info. Java Garbage Collection seems to be working fine as
always. Rebooting the services doesn't free the network sockets and it
continues to be a problem until the server is rebooted.
1. Use Java Garbage Collection
Modify the JVM.CONFIG file and and append the following
with fusion reactor u can store request data which may help when it is
totally unresposnive
On Thu, Sep 22, 2011 at 9:46 PM, Richard Steele r...@photoeye.com wrote:
Thanks for the info. Java Garbage Collection seems to be working fine as
always. Rebooting the services doesn't free the
I am having the same problem.. Sometimes restarting the cold fusion
service and/or web publishing service fixes it but most of the time
doesn't. I saw something really strange in the performance monitor
that I can't understand. There is only 1 website running on the
server - with usually very
there are quite a number of things that can cause high number of
user/connections.
bots, webcrawlers are most common.
cf will regulalry get brute force attacks on the cf/administrator or RDS
your event logs and weblogs should give u some idea
On Thu, Sep 22, 2011 at 11:34 PM, Al Musella, DPM
We're starting to get CF8 Enterprise hung instances every 3=4 days or so at
different times. For no apparent reason, requests don't complete on two load
balanced instances. Restarting both instances does not help. The instances
become hung again very quickly. Restarting IIS for these instances
We're starting to get CF8 Enterprise hung instances
every 3=4 days or so at different times.
Hi Rick, I'd be looking at the memory usage. A tool such as
FusionReactor or SeeFusion can help track down problems pretty
quickly. In many cases, this kind of behavior points to improperly
scoped
15 matches
Mail list logo
