RE: [fusionreactor] FusionReactor all of a sudden stops listening to Railo Instance
Charlie Arehart
couple of things.
First, I notice that the log you refer to below both FRAM and your "Railo
instance". And you refer to port 8088. Just to be clear, the FRAM instance
would be one thing (running as a standalone Tomcat instance, though
identified in your Windows processes as and Service definition as
"FusionReactor AM Service"), and it would "listen" on port 8087. That would
present an interface (looking like the normal FR interface) by which you can
manage other FR instances.
Then you would have the FR instance within your Railo deployment, and that
would typically (if it's the only FR instance you have) be "listening" on
port 8088. Now, by listening I mean that if you requested in your browser
that machine name or some given site you've defined in IIS or Apache, and
port 8088, you'd then see the FR interface for that instance. And perhaps
that's what you mean by saying you don't find it working. But what if you
remove the port? What if you say, for instance,
http://yourmachineorsitename/fusionreactor/fhtml.cfm? Does that work? It
might. If it does, then it shows that FR is working and accessible.
As for port 8088, that's an internal web server that FR enables within
itself (within the Railo or CF instance) as a separately accessible web
server (yep, different from any external web server, like IIS or Apache, and
different from the internal web server in CF, or the Tomcat web server if
that's where you've deployed Railo on a standalone Tomcat install.) But as
I'm showing in the last paragraph, you don't HAVE to use that built-in FR
port. You can, by default, use an external web server to talk to it, if that
external web server/site is configured to pass .cfm files to CF (or Railo,
in your case). If that surprises any readers, you can turn that off (so that
FR is NOT accessible through an external web server).
Finally, you saying the last paragraph, "Here is the reactor.log from the
Railo instance. You can see that after 33 minutes, it just shuts down. Now,
keep in mind that Railo/Tomcat itself did NOT go down. FRAM just stopped
listening. During this time, I am still able to access FRAM".
First, you refer to this as being the reactor.log from the Railo instance,
but then you refer to "FRAM just stopped listening" and yet "I am still able
to access FRAM". I want to make sure you really mean to be referring to FRAM
(at port 8087) or the FR instance within Railo (which I assume would be at
8088). They're not the same, so can you clarify?
Second, the log entries that you show referring to "stopping" and "shutdown"
are not about FR itself, but about the "Federated Data manager" and
FA-Connect. The former is (I think) about the FRAM instance managing the
logs for the other FR instances, and the latter refers to FA,
FusionAnalytics, and would seem to be related to any connection you've made
from the instance to an implementation you have somewhere of
FusionAnalytics. Do you have one? And even if so, I don't think this
"shutdown" of that connection has anything at all to do with your FR
instance responding itself.
I realize you may just be grasping at straws if somehow you are finding that
you can't access FusionReactor. And I'm sure the FR guys will reply as well
and may have other things to say. Since it's later in the day and they are
in Europe, they may not necessarily reply today, so I wanted to offer some
thoughts in the meantime. Also, if you may respond to me with
clarifications, that may help them in helping you more specifically. Hope
so.
/charlie
-----Original Message-----
From: fusion...@googlegroups.com [mailto:fusion...@googlegroups.com]
On Behalf Of Jason101
Sent: Friday, March 29, 2013 2:21 PM
To: fusion...@googlegroups.com
Subject: [fusionreactor] FusionReactor all of a sudden stops listening to
Railo Instance
Hello,
I'm having an issue with FusionReactor all of a sudden just stops listening
to my Railo Instance. When FusionRecator works, it usually one works for a
few hours, and then the connection to port 8088 just gets dropped, and the
only way to get it back is to restart Tomcat. Even then, I have to restart
Tomcat 2 or 3 times before FusionReactor will start listening to that
instance. FusionReactor for some reason won't start listening on the first
restart of Tomcat. I even gave it 5 or so minutes to see if it takes a
while.
I only have 1 instance installed for Railo (C:\Railo\) I did a netstat and I
don't see 8088. I have also confirmed that there are no conflicts that
nothing else is listening on that port.
Here are the versions I am using
OS: Windows Server 2012 x64
Ram: 32GB
FusionReactor: Revision: 4.5.3, Build: fusionreactor-4.5.140.29184
Railo: Railo 4.0.4.001 final
Tomcat: Apache Tomcat/7.0.33
Here is the reactor.log from the Railo instance. You can see that after 33
minutes, it just shuts down. Now, keep in mind that Railo/Tomcat itself did
NOT go down. FRAM just stopped listening. During this time, I am still able
to access FRAM
Its strange that even though Fusionreactor was not listening anymore, there
was still data being recorded to JDBC.log (but not any others)
2013-03-28 23:00:00.049 INFO Setting log level INFO
2013-03-28 23:00:00.959 INFO Archive: Age Retention Strategy: Purging
archive (limit 31 days, or: Mon Feb 25 00:00:00 EST 2013)
2013-03-28 23:00:00.963 INFO Archive: Age Retention Strategy: The oldest
folder (20130327) is still newer than the purge limit, which is (20130225).
Purge complete.
2013-03-28 23:33:11.020 INFO FederatedDataManager: Signalling stop to all
workers.
2013-03-28 23:33:11.020 INFO FusionReactor Enterprise Service - Local
Metrics Cache - Halting.
2013-03-28 23:33:11.021 INFO FederatedDataManager: Signal complete
2013-03-28 23:33:11.021 INFO FederatedDataManager: FusionReactor FD/R Remote
DataCollector for {Railo.WaveWorks-Web1} - Halted.
2013-03-28 23:33:11.021 INFO FACScheduler: Shutting down scheduler.
2013-03-28 23:33:11.023 INFO FA-Connect: Stopping workers.
2013-03-28 23:33:11.023 INFO FA-Connect: FusionReactor FusionAnalytics
Connector Worker 0: Signal halt.
2013-03-28 23:33:11.523 INFO FA-Connect: FusionReactor FusionAnalytics
Connector Worker 1: Signal halt.
2013-03-28 23:33:12.024 INFO FA-Connect: FusionReactor FusionAnalytics
Connector Worker 2: Signal halt.
2013-03-28 23:33:12.525 INFO FA-Connect: FusionReactor FusionAnalytics
Connector Worker 3: Signal halt.
2013-03-28 23:33:13.026 INFO FA-Connect: Shutdown complete.
I only have 4 days left on my trial, and would love to get this resolved
prior to obtaining a licensing.
Thanks!
--
You received this message because you are subscribed to the Google Groups
"FusionReactor" group.
To unsubscribe from this group and stop receiving emails from it, send an
email to fusionreacto...@googlegroups.com.
To post to this group, send email to fusion...@googlegroups.com.
Visit this group at http://groups.google.com/group/fusionreactor?hl=en.
For more options, visit https://groups.google.com/groups/opt_out.
Jason101
Thanks Charlie,
I played around with this for another two hours tonight, and unfortunately no results. When I mentioned FRAM, I meant at port 8087. I am always able to access FRAM at port 8087. However, I can't access the Railo instance at 8088. I also confirmed I was not able to access the railo instance from IIS (without entering port 8088). It doesn't appear that mapping is there as its giving me a 404 like its looking for a local "fusionreactor" directory relative to the website and not a mapping (like that of "railo-context") I was able to access port 8088 for about 2 minutes, and then I could no longer access it. I don't see anything written to the logs in the instance log folder (except for the jdbc.log I mentioned in my last post)
David Stockton
Jason101
After working with the FusionReactor team, they determined my problem is a result frequent re-deploys of my web applications.
“Looking at your logs, you appear to have a problem with your Tomcat/Railo installation which is causing frequent re-deploys of your web applications. Quite commonly, this is if you have a process which is touching your web.xml (or similar) files. When Tomcat detects the file "touch" it will re-deploy the web-application. If this happens rapidly or frequently FusionReactor won't come back up. FusionReactor v4 does not support live re-deployment of Java web-applications”
They recommended I figure out what’s causing the redeployment which should solve remedy why FusionReactor is disappearing after a short space of time. I’m thinking this is a result of mod_cfml-valve which is used on on Windows Railo installs which dynamically re-created contexts when Tomcat is restarted and when passed from Boncode Connetor/IIS for the first time.
I was going to disable mod_cfml-valve and just create the contexts manually, but unfortunately, my trial of FusionReactor has expired, and I can no longer pursue the issue. And I can’t in good conscience purchase a license knowing it may not work.
FusionReactor Trial on our test box, which only runs 2 contexts/applications does not have this issue, and it FusionReactor works flawlessly, so I am leaning towards the 20+ contexts that need to be created on the fly by mod_cfml-valve as the contributing issue.
Has
anyone running FusionReactor on Windows running Railo had a similar issue?