* Log Analyzer code has been completely refactored resulting in better performance and stability and more maintainable code.
* The progress indicator for downloads in Log Analyzer page is now a true quantitative progress bar.
* Error messages in Log Analyzer were rewritten. Every single possible error is now differentiated by a specific message providing help for the situation.
* “Group by” option in history/trends page could deliver incorrect results if MySQL and MONyog were using different timezones.
* With SFTP connection to the log file selecting “ALL” for log size to be downloaded in Log Analyzer page could have the result that the log file download would last forever.
* Analyzing log chunks containing BULK INSERT statements could in special situations cause a program crash.
* Fixed a bug with resolvement of time in log analyzer that could result that no queries or less queries than what should be were returned. The most recent queries would sometimes not be shown. Issue was only if log was retrieved with SFTP.
* When retrieving logs stored in tables the substitution of literals would fail for binary data.
* When downloading a log MONyog would continue to download in the background until end of file even after STOP button was clicked.
* In the display of queries from the Processlist page special characters could garble.
* Also in Log Analyzer page special characters would not always be indentified correctly. This bug also affected records with binary data. Both display and grouping was affected (depending on where those occurred – in plain strings or in identifiers).
* Repeatedly switching between sorting and show details for a slow query log in log analyzer could in certain situations crash MONyog.
* Installing a RPM multi/single server build on top of a trial build (and vice versa) resulted in that they were installed in parallel. Also on SuSE MONyog was not registered with the correct runlevel. The RPM has been updated to fix this.
* Display of servers is now sorted in the order of the server/connection name. Earlier the internal ID was used for sorting. Also note that now no MONyog restart is required for changing sort order. A connection rename alone will do.
As compared to the 2.6x betas this also includes:
* Reporting the execution time of a query in 'query sniffer' was inaccurate. This was an issue with the 2.6 betas only. 2.5x were not affected.