web stats
Mirth Community - View Single Post - Intermittently can't launch Administrator with 'Error Downloading/Verifying Codebase'
View Single Post
  #1  
Old 12-31-2018, 11:07 AM
jridderhoff jridderhoff is offline
OBX.1 Kenobi
 
Join Date: Jan 2015
Posts: 42
jridderhoff is on a distinguished road
Default Intermittently can't launch Administrator with 'Error Downloading/Verifying Codebase'

We've been encountering a frustratingly sneaky problem recently when attempting to connect to our instances of Mirth Connect 3.6.1. Sometimes it works fine, but seemingly at random for extended periods of time, any attempt to connect the launcher (either via the NextGen Connect Administrator Launcher or via the Mirth Connect Administrator web page) fails.

When attempted via the web page, it just times out; when attempting via the Administrator Launcher (using the built-in JVM), we get the following error:

Code:
Error Downloading/Verifying Codebase
The markup in the document preceding the root element must be well-formed.
org.xml.sax.SAXParseException; lineNumber: 1; columnNumber: 3; The markup in the document preceding the root element must be well-formed.
	at com.sun.org.apache.xerces.internal.parsers.DOMParser.parse(Unknown Source)
	at com.sun.org.apache.xerces.internal.jaxp.DocumentBuilderImpl.parse(Unknown Source)
	at javax.xml.parsers.DocumentBuilder.parse(Unknown Source)
	at com.mirth.connect.client.launcher.f.a(SourceFile:664)
	at com.mirth.connect.client.launcher.f.a(SourceFile:167)
	at com.mirth.connect.client.launcher.MirthClientLauncher.run(SourceFile:1122)
	at java.lang.Thread.run(Unknown Source)
During this time, we can still get to the Administrator Launcher web page in a browser (//<host>:<http.port>/), but attempting to actually click the "Launch Mirth Connect Administrator" button just times out.

There are no errors logged in mirth.log, and we know the port is open because we can still telnet to the host using the port configured as http.port in mirth.properties. Running ps -ef | grep mcservice still shows that mcservice is running. There are other applications running on the same server, and there is ample available RAM and CPU%.

Even more oddly, the problem seems to self-resolve after an extended period of time. Coming back and attempting again some hours later, it's as if there was never a problem (nothing logged, everything responding as expected).

We're running:
CentOS Linux release 7.6.1810 (Core)
Oracle jdk1.8.0_191-amd64
Mirth Connect 3.6.1
NextGen Connect Administrator Launcher using Default Java Home (1.8.0_191-b12)

Any thoughts or further troubleshooting suggestions? At the moment, if we urgently need to access the Administrator console during one of these recalcitrant periods, our only option is to kill the mcservice process and restart the service again (usually in this phase, running mcservice stop doesn't actually the process, despite feedback saying that the daemon has been stopped).

Thanks!
Reply With Quote