I am using the standard environment of GAE / PY.
I'd like to confirm syslog, but I could not confirm it with stackdriver logging(only request_log and activity_log).
Can not you confirm syslog with App Engine standard environment?
You do not have access to the actual syslog of the systems on top of which your standard env GAE app instances are running. I suspect the user-provided app code might not even have write permissions to such logfile.
If you're looking for logs produced by your app (for example by using the logging facility, as recommended by ), they're bundled under the request_log entries for the requests that triggered the respective code execution, see Reading Application Logs on Google App Engine from Developer Console
Related
Liberty has rich support for logging and tracing as outlined in https://www.ibm.com/support/knowledgecenter/SSEQTP_liberty/com.ibm.websphere.wlp.doc/ae/rwlp_logging.html
Based on that, I am able to get audit, messaging, and trace logs sent to the console. However, one part of our system is using Liberty's JBatch implementation and these logs are getting lost inside our containers' ephemeral filesystem.
Is there a way to force Liberty to send these logs to the console with the rest of our output (so that it will get shipped to our log analysis service)?
Apologies in advance for somewhat vague information. I am new to Citrix XenApp/XenDesktop technology and am just looking for generic troubleshooting information.
At my place of employment we have kiosks that are configured to connect to a SaaS webapp. These kiosk have either the Citrix XenApp or XenDesktop installed.
One of the icons launches the IE browser that connects to the SaaS app using a preconfigured user account. Sometimes, however, instead of launching the browser, the system displays the "The session limit has been reached. Please contact your system administrator." error shown in below image.
The people administering these kiosks think that this message comes from the SaaS web application but that application does not enforce any limits on how many session are open for a given account under a given time.
Also considering how Citrix XenApp/XenDesktop works I would think (but maybe I am wrong) that if the SaaS app did reject a user login, we would be displayed an error message in Internet Explorer instead of this ICA prompt.
So I think that the issue here could be that the message is not about login sessions made to the background SaaS app but either about Citrix sessions or perhaps previous IE browsers somehow running in the background(?)
However our company's Citrix team looked at this and noticed that "Citrix was still active" when this prompt was displayed. The conclusion was then that Citrix is for that reason not the cause here.
So I wanted here to ask some questions on what things I could consider as causes and where I could look in the hopes of getting started on this issue.
This would be for XenApp / XenDesktop 7.18.
The questions I have:
Does XenApp / XenDesktop have log files that can be consulted for
debugging issues like this?
Is it possible to get XenApp / XenDesktop to run in debug mode (to
output more details to the log files)?
Does Citrix have configuration settings that could lead it to
have an issue like this?
A. First check the event logs and see when you facing this issue so does any event logs generated.
B. Also you can check the ICA configuration tool for session settings and checked if session settings are set to NEVER.
C. The ICA listener configuration tool is located at Start > All Programs > Citrix > Administration Tools > ICA Listener Configuration.
You are on the right track with the SaaS application itself reporting the error. If this Citrix session was already active when the icon was clicked again and the preconfigured user was already logged into the SaaS application, that would account for this error. To investigate, logout the Citrix session and try clicking the icon again, or check SaaS application to see if that preconfigured users is already connected.
Is the same user used for all these kiosks or is each kiosk supposed to have a unique user? Can this preconfigured user log in multiple times?
I'm building a hybrid app that speaks to a Bluemix app. I've got Mobile Application Security and Mobile Quality Assurance as a service. I'm using IBMLogger to send log messages, like so:
IBMBluemix.getLogger().info("Device successfully registered: "+ JSON.stringify(response));
I was under the impression that the MobileFirst-style services on Bluemix acted somewhat like the same components for a local MF install. I'm trying to find where I can see my log files in the BM console but have not been able to find them.
For a local MF install using the hybrid SDK, you have to specifically say you want your logs sent to the server, but I do not see that option when looking at the hybrid docs for BM/MF.
There is some doc here: https://www.ng.bluemix.net/docs/starters/mobile/mobilecloud/nodejsmobile.html#log
I suppose by default the log goes to the console, which you can retrieve through the loggregator, aka do cf logs <yourapp> --recent. You can also use the Monitoring & Analytics service to retain logs - see the above doc.
We have a handful of Ruby (Rails/Sinatra) apps and are looking for an easy means of managing retention, search and analysis of our logs for these applications.
The initial problem was that every time we'd push a new version of our apps the logs would disappear.
We then started streaming our logs to a file via a terminal using:
cf logs AppName
however the logfiles get very big, very fast and quickly become a problem.
We know that the Bluemix Monitoring and Analytics service provides a lot of the function we need. We're looking that over but want to know if there are other recomended/proven options.
Thanks
We found several 3rd party apps that provide the functions we need.
To use any of these we first had to configure third party logging on Bluemix and used the steps below.
Any 3rd party logger that supports the syslog protocol can be used. The initial setup, registration and configuration of the log manegement service, is well covered at Configuring Selected Third-Party Log Management Services.
What will come out of the configuration step is a syslog URL which will be the destination for your logs.
Once the logging service is configured a user-provided service instance needs to be created to stream the logs to the logging service. We did this using:
cf create-user-provided-service <user-provided-service_name> -l <syslog_URL>
Last step is to bind the service instance to our Ruby apps.
cf bind-service AppName <user-provided-service_name>
For the changes to take effcet, we then had to restage our ruby apps:
cf restage AppName
There was a brief delay between when we'd see the logs generated and when they'd show up in the logger service but overall this is working out ok for us so far.
Any good suggestion for Monitoring JBOSS 7 in Production ? I would also like to configure alerts based on certain condition. Of course , It has to be Open source.
Thanks.
You can use standard JConsole that comes with JBoss dependencies added. It's used to monitor your servers state and mbeans, it's very useful.
To test it on localhost start your server and then run the JConsole from your server/bin directory and select JBoss in the Local process selection.
To use it on "remote" server, start your server on "REMOTE_HOST" and then run JConsole from an JBoss/bin directory and connect with the followin string
service:jmx:jmx-remoting://REMOTE_HOST_NAME:9999 (or the port you use) and enter the username and password.
Secondly, for more detailed info of objects creation, memory leaks, CPU% (profiling) there is another one as:
http://jbossprofiler.jboss.org/
You can try to use free open source APM like scouter.
It shows very useful realtime performance information of every request.
And also you can set a threshold of resources and can make plugin for alerting to external.
https://github.com/scouter-project/scouter
JBoss7 need to set module option.
-Djboss.modules.system.pkgs=~~~,scouter