CFChart stopped working with ColdFusion (2016 release) Update 10 (release date, 01 March 2019)
I have an issue with CFCHART that I am hoping to get resolved. Back in February I noticed that our cfchart will not produce a chart if the format = png or jpg. Only html format would produce a chart. I came to forums looking for a solution and found this thread (CFCHARTformat=png image not loading on browser ). I had our server admin change the cf chart cache type setting from disk to memory and then voila! - the png and jpg charts started to work as expected. After this most recent release/update it seems the CFCHART is no longer working. After some minor debugging I found that we are back at square one with charts only being generated if the format is HTML. I was thinking maybe changing the cache back to disk and attempt to check permissions to make sure nothing is blocking the CFFILESERVLET Mapping but the issue is that I do no have access to the server/server settings. I will have to communicate with a server admin on the exact steps to take to check these permissions/settings. I was wondering if anyone has dealt with this specific issue before? Any insight/advice would be helpful. Thanks!
UPDATE - HERE IS ERROR LOG - HOPEFULLY THIS HELPS
Mar 26, 2019 11:09:05 AM org.apache.catalina.core.StandardWrapperValve invoke
SEVERE: Servlet.service() for servlet [CfmServlet] in context with path [] threw exception [ROOT CAUSE:
java.lang.OutOfMemoryError: Metaspace
] with root cause
javax.servlet.ServletException: ROOT CAUSE:
java.lang.OutOfMemoryError: Metaspace
ISSUE RESOLVED -
We increased the Metaspace memory and it resolved this issue. This may be an internal bug with Coldfusion. I am very new to Coldfusion development but it seems that this issue should not occur. My biggest concern is changing the setting in CF Administrator settings(under charting) from disk to memory cache. Technically, according to the description of the page, the memory cache should be cleaning/clearing itself. If that is not working then there could be a possible memory leak. I will continue to monitor the situation and update this post with any subsequent findings.
