14:00:25 #startmeeting monasca 14:00:26 Meeting started Wed Jun 7 14:00:25 2017 UTC and is due to finish in 60 minutes. The chair is rhochmuth. Information about MeetBot at http://wiki.debian.org/MeetBot. 14:00:27 Useful Commands: #action #agreed #help #info #idea #link #topic #startvote. 14:00:29 The meeting name has been set to 'monasca' 14:01:10 uh oh, no agenda for today's meeting 14:01:19 that's a first 14:01:21 o/ 14:01:33 anyone around today? 14:01:36 o/ 14:01:38 o/ 14:01:49 o/ 14:02:01 hi koji and hosanai and haruki 14:02:07 o/ 14:02:11 hello 14:02:24 hi 14:02:40 hello 14:02:43 i guess the new time-slot works out a little better for everyone in japan 14:02:51 but, it is still a little late 14:03:22 yeah, it works for me :-) 14:03:27 11:00pm on our time zone now 14:03:34 does anyone have any topics or reviews 14:03:39 my wife is in japan 14:03:47 she called me at 4:00 AM 14:04:05 hahaha. she tries to kill you :-) 14:04:14 she does her best 14:04:34 I would like to share info of analytics status. 14:04:35 torture is her specialty though 14:04:50 hosanai: you have the floor 14:04:55 #topic monasca-analytics 14:05:00 thanks! 14:05:12 brad gave us data for analytics 14:05:12 interesting! 14:05:27 brad klein? 14:05:37 yeah 14:05:54 very nice 14:06:02 and Fdaisuke checked the data but information is not enough for analytics 14:06:22 what data did you get? 14:06:35 We got it. 14:06:50 i meant type of data 14:06:55 was it metrics, logs? 14:07:20 sorry, i miss understood. Fdaisuke, could you explain it? 14:08:28 ok 14:09:41 there is no metorisc and logs, just only "timestamps" and "value" (100% or not) 14:10:21 my rough understanding is we got timestamp, value and value_meta for each services such as swift, nova and so on 14:10:52 do you have the metric name and dimensions? 14:10:55 thanks, hosanai 14:11:30 isn't the "value" a metric? 14:12:52 memory usage or like that? 14:13:33 in dimensions, there is service and in measurements we have timestamp, value and value_meta. i think. 14:14:13 it looks status of services. 14:14:25 and are there any points in which incidents have been labeled? 14:15:10 for example, do you have a set of timestamps and label, for an incident and/or type of incident 14:15:22 yeah, he explained us unexpected behavior and the time. 14:16:12 so, you have a bunch of metrics, and then a set of incidents and the timestamps for when they occured it sounds like 14:17:04 we don't have metric like memory in the data. 14:17:16 only service status 14:18:08 s/memory/memory usage/ 14:18:49 oh, so basically if swift goes down, the value of some metric, such as status with dimension service=swift is 1.0 14:19:45 only 1 or 0? 14:20:56 Fdaisuke, could you give us an example? you said the data looks like 100%, 90%... 14:22:31 i saw 100%, 90%, 40%, sometimes 0% 14:23:04 ok, so other than the status metric, is there anything else of interest? 14:23:28 if that is all you have, then i agree, the dataset is not going to offer a lot of insight 14:23:45 yeah so Fdaisuke sent email to brad to get additional data. 14:23:48 or data analysis opportunites 14:24:19 i think what you need is a lot of metrics 14:24:36 the status metric is already informing you about incidents 14:24:54 i think so. I hope brad keep the data :-) 14:25:37 if you had more data, you could potentially identity other metrics that represent the incident 14:25:52 or possibly RCA 14:26:15 hopefully :-) 14:26:45 it would also be interesting to have the set of alarms that occurred 14:27:19 then you could train your data set using the alarms as the response/labels 14:27:49 and then try and predict incidents when you don't have the alarms 14:28:39 i see the alarms are useful to create labels. 14:28:43 in other works can you develop a model for predicting alarms 14:29:03 yeah, i think that would be an interesting problem 14:29:37 given a set X of all the metrics and a set y that consists of all the alarms 14:30:10 brad explained us if the status is not 100%, it's a problem from operator point of view. 14:30:35 so, status can be used too 14:30:50 i guess 0 is no problem 14:31:15 25% could be low status 14:31:23 100% could be there is a problem 14:31:34 > 75% could be there is a high probability there is a problem 14:31:46 so, that could be useful if you had the other metrics 14:31:55 you could use status as the response/label set 14:32:19 and then train on that using the metrics as predictors 14:32:30 thanks for the info/advise:-) 14:32:33 similar to how alarms would be used 14:32:44 your welcom 14:32:50 that's all I would like to share 14:32:51 i think you need a lot more data 14:33:06 yeah that's right 14:33:24 you could also take the status variables, and treat it as binary 14:33:35 for example, anything above 50% is a problem 14:33:54 then your problem turns into a binary classification problem 14:34:03 but, i'm not sure that is what you want to do 14:34:28 a binary classification means 2 values classification? 14:34:34 correct 14:35:06 but, you can also treat the predictor as continous, which is what the status is 14:36:14 it's a good way to handle the problem as a binary classification to get good result 14:36:15 so, if brad can get you more data, even without the alarms, i think you might have something you can start with 14:37:13 yeah, we can push analytics ahead with brad's help 14:37:23 it would also be very interesting to know what metrics correlate well with problems 14:38:03 i think so it's kind of RC 14:38:38 that would be good for RCA 14:40:42 yeah, and witec introduced Fujitsu's colleagues who created original algorithm (i forgot the approach) for analytics 14:41:10 do you know the names? 14:42:05 i forgot it I got a paper but did't read it yet :( 14:44:02 names for the guys? 14:44:10 you mean 14:44:32 well, i was wondering if they were hpe or fujitsu folks 14:46:09 I will send you the names later. I can't access company email now. 14:46:26 np 14:46:49 at least 9 hours later :-) 14:47:29 i thought you were still at work 14:47:33 i'm dissapointed 14:47:38 :-) 14:47:50 hahaha 14:49:45 so, is that it for today? 14:50:46 i don't have anything now 14:51:02 thx hosanai 14:51:13 i guess i'll end the meetign a little early 14:51:29 #endmeeting