- Looking into Visual Editor validation exceptions. TL;DR; is that the rate of errors per valid events is about 0.5%, so nothing to worry about.
milimetric@eventlog1001:/srv/log/upstart$ ls -l eventlogging_processor-client-side-events.log*
-rw-r----- 1 root root 16499196 May 28 23:59 eventlogging_processor-client-side-events.log
-rw-r----- 1 root root 2148452 May 28 06:35 eventlogging_processor-client-side-events.log.1.gz
-rw-r----- 1 root root 2235029 May 27 06:36 eventlogging_processor-client-side-events.log.2.gz
-rw-r----- 1 root root 2165733 May 26 06:36 eventlogging_processor-client-side-events.log.3.gz
milimetric@eventlog1001:/srv/log/upstart$ ionice -c3 sudo zcat eventlogging_processor-client-side-events.log.1.gz | ~/count.invalid.from.processor.log.py
defaultdict(<type 'int'>, {u'saveIntent': 22, u'saveFailure': 37, u'abort': 740, u'saveSuccess': 33, u'ready': 178, u'saveAttempt': 7})
milimetric@eventlog1001:/srv/log/upstart$ ionice -c3 sudo cat eventlogging_processor-client-side-events.log | ~/count.invalid.from.processor.log.py
defaultdict(<type 'int'>, {u'saveIntent': 21, u'saveFailure': 47, u'init': 2, u'abort': 634, u'saveSuccess': 38, u'ready': 129, u'saveAttempt': 18})
milimetric@eventlog1001:/srv/log/upstart$ ionice -c3 sudo zcat eventlogging_processor-client-side-events.log.2.gz | ~/count.invalid.from.processor.log.py
defaultdict(<type 'int'>, {u'saveIntent': 16, u'saveFailure': 79, u'init': 1, u'abort': 785, u'saveSuccess': 38, u'ready': 151, u'saveAttempt': 18})
milimetric@eventlog1001:/srv/log/upstart$ ionice -c3 sudo zcat eventlogging_processor-client-side-events.log.3.gz | ~/count.invalid.from.processor.log.py
defaultdict(<type 'int'>, {u'saveIntent': 20, u'saveFailure': 28, u'init': 4, u'abort': 713, u'saveSuccess': 28, u'ready': 179, u'saveAttempt': 7})
In summary, for Visual Editor Event Logging,
May 28th: 889 errors, 185522 valid events
May 27th: 1017 errors, 205462 valid events
May 26th: 1088 errors, 216707 valid events
May 25th: 979 errors, 190894 valid events
- Talked to Chris Steipp, will reconsider the DNT discussion with his point of view. By next week we should have a rough first version of the high level requirements. We'll take this around to help get estimates and iterate with stakeholders to see if the timeline is acceptable.
