Bearloga has uploaded a new change for review. ( https://gerrit.wikimedia.org/r/341742 )
Change subject: Annotate Reportupdater migration on graphs ...................................................................... Annotate Reportupdater migration on graphs Change-Id: Ie650b1eb0f5c9cc40e43a316b71f44e0b8b8cab7 --- M server.R M tab_documentation/app_events.md M tab_documentation/app_load.md M tab_documentation/click_position.md M tab_documentation/desktop_events.md M tab_documentation/desktop_load.md M tab_documentation/failure_breakdown.md M tab_documentation/failure_langproj.md M tab_documentation/failure_rate.md M tab_documentation/failure_suggests.md M tab_documentation/fulltext_basic.md M tab_documentation/geo_basic.md M tab_documentation/invoke_source.md M tab_documentation/kpi_api_usage.md M tab_documentation/kpi_augmented_clickthroughs.md M tab_documentation/kpi_load_time.md M tab_documentation/kpi_zero_results.md M tab_documentation/language_basic.md M tab_documentation/mobile_events.md M tab_documentation/mobile_load.md M tab_documentation/open_basic.md M tab_documentation/paulscore_approx.html M tab_documentation/prefix_basic.md M tab_documentation/survival.md 24 files changed, 100 insertions(+), 57 deletions(-) git pull ssh://gerrit.wikimedia.org:29418/wikimedia/discovery/rainbow refs/changes/42/341742/1 diff --git a/server.R b/server.R index 343dd5f..158ecf5 100644 --- a/server.R +++ b/server.R @@ -69,7 +69,8 @@ polloi::smoother(smooth_level = polloi::smooth_switch(input$smoothing_global, input$smoothing_desktop_event)) %>% polloi::make_dygraph(xlab = "Date", ylab = "Events", title = "Desktop search events, by day") %>% dyRangeSelector %>% - dyEvent(as.Date("2016-07-12"), "A (schema switch)", labelLoc = "bottom") + dyEvent(as.Date("2016-07-12"), "A (schema switch)", labelLoc = "bottom") %>% + dyEvent(as.Date("2017-01-01"), "R (reportupdater)", labelLoc = "bottom") }) output$desktop_load_plot <- renderDygraph({ @@ -77,7 +78,8 @@ polloi::smoother(smooth_level = polloi::smooth_switch(input$smoothing_global, input$smoothing_desktop_load)) %>% polloi::make_dygraph(xlab = "Date", ylab = "Load time (ms)", title = "Desktop load times, by day", use_si = FALSE) %>% dyRangeSelector %>% - dyEvent(as.Date("2016-07-12"), "A (schema switch)", labelLoc = "bottom") + dyEvent(as.Date("2016-07-12"), "A (schema switch)", labelLoc = "bottom") %>% + dyEvent(as.Date("2017-01-01"), "R (reportupdater)", labelLoc = "bottom") }) output$paulscore_approx_plot_fulltext <- renderDygraph({ @@ -149,14 +151,16 @@ mobile_dygraph_set %>% polloi::smoother(smooth_level = polloi::smooth_switch(input$smoothing_global, input$smoothing_mobile_event)) %>% polloi::make_dygraph(xlab = "Date", ylab = "Events", title = "Mobile search events, by day") %>% - dyRangeSelector + dyRangeSelector %>% + dyEvent(as.Date("2017-01-01"), "R (reportupdater)", labelLoc = "bottom") }) output$mobile_load_plot <- renderDygraph({ mobile_load_data %>% polloi::smoother(smooth_level = polloi::smooth_switch(input$smoothing_global, input$smoothing_mobile_load)) %>% polloi::make_dygraph(xlab = "Date", ylab = "Load time (ms)", title = "Mobile search events, by day", use_si = FALSE) %>% - dyRangeSelector + dyRangeSelector %>% + dyEvent(as.Date("2017-01-01"), "R (reportupdater)", labelLoc = "bottom") }) ## App value boxes @@ -192,28 +196,32 @@ android_dygraph_set %>% polloi::smoother(smooth_level = polloi::smooth_switch(input$smoothing_global, input$smoothing_app_event)) %>% polloi::make_dygraph(xlab = "Date", ylab = "Events", title = "Android mobile app search events, by day") %>% - dyRangeSelector + dyRangeSelector %>% + dyEvent(as.Date("2017-01-01"), "R (reportupdater)", labelLoc = "bottom") }) output$android_load_plot <- renderDygraph({ android_load_data %>% polloi::smoother(smooth_level = polloi::smooth_switch(input$smoothing_global, input$smoothing_app_load)) %>% polloi::make_dygraph(xlab = "Date", ylab = "Load time (ms)", title = "Android result load times, by day", use_si = FALSE) %>% - dyRangeSelector + dyRangeSelector %>% + dyEvent(as.Date("2017-01-01"), "R (reportupdater)", labelLoc = "bottom") }) output$ios_event_plot <- renderDygraph({ ios_dygraph_set %>% polloi::smoother(smooth_level = polloi::smooth_switch(input$smoothing_global, input$smoothing_app_event)) %>% polloi::make_dygraph(xlab = "Date", ylab = "Events", title = "iOS mobile app search events, by day") %>% - dyRangeSelector + dyRangeSelector %>% + dyEvent(as.Date("2017-01-01"), "R (reportupdater)", labelLoc = "bottom") }) output$ios_load_plot <- renderDygraph({ ios_load_data %>% polloi::smoother(smooth_level = polloi::smooth_switch(input$smoothing_global, input$smoothing_app_load)) %>% polloi::make_dygraph(xlab = "Date", ylab = "Load time (ms)", title = "iOS result load times, by day", use_si = FALSE) %>% - dyRangeSelector + dyRangeSelector %>% + dyEvent(as.Date("2017-01-01"), "R (reportupdater)", labelLoc = "bottom") }) output$click_position_plot <- renderDygraph({ @@ -235,7 +243,8 @@ dyAxis("x", ticker = "Dygraph.dateTicker", axisLabelFormatter = polloi::custom_axis_formatter, axisLabelWidth = 100, pixelsPerLabel = 80) %>% dyLegend(labelsDiv = "app_invoke_source_legend") %>% - dyRangeSelector(fillColor = "", strokeColor = "") + dyRangeSelector(fillColor = "", strokeColor = "") %>% + dyEvent(as.Date("2017-01-01"), "R (reportupdater)", labelLoc = "bottom") }) ## API plots @@ -243,35 +252,40 @@ split_dataset$cirrus %>% polloi::smoother(smooth_level = polloi::smooth_switch(input$smoothing_global, input$smoothing_fulltext_search)) %>% polloi::make_dygraph(xlab = "Date", ylab = "Searches", title = "Full-text via API usage by day", legend_name = "Searches") %>% - dyRangeSelector + dyRangeSelector %>% + dyEvent(as.Date("2017-01-01"), "R (reportupdater)", labelLoc = "bottom") }) output$open_aggregate <- renderDygraph({ split_dataset$open %>% polloi::smoother(smooth_level = polloi::smooth_switch(input$smoothing_global, input$smoothing_open_search)) %>% polloi::make_dygraph(xlab = "Date", ylab = "Searches", title = "OpenSearch API usage by day", legend_name = "Searches") %>% - dyRangeSelector + dyRangeSelector %>% + dyEvent(as.Date("2017-01-01"), "R (reportupdater)", labelLoc = "bottom") }) output$geo_aggregate <- renderDygraph({ split_dataset$geo %>% polloi::smoother(smooth_level = polloi::smooth_switch(input$smoothing_global, input$smoothing_geo_search)) %>% polloi::make_dygraph(xlab = "Date", ylab = "Searches", title = "Geo Search API usage by day", legend_name = "Searches") %>% - dyRangeSelector + dyRangeSelector %>% + dyEvent(as.Date("2017-01-01"), "R (reportupdater)", labelLoc = "bottom") }) output$language_aggregate <- renderDygraph({ split_dataset$language %>% polloi::smoother(smooth_level = polloi::smooth_switch(input$smoothing_global, input$smoothing_language_search)) %>% polloi::make_dygraph(xlab = "Date", ylab = "Searches", title = "Language Search API usage by day", legend_name = "Searches") %>% - dyRangeSelector + dyRangeSelector %>% + dyEvent(as.Date("2017-01-01"), "R (reportupdater)", labelLoc = "bottom") }) output$prefix_aggregate <- renderDygraph({ split_dataset$prefix %>% polloi::smoother(smooth_level = polloi::smooth_switch(input$smoothing_global, input$smoothing_prefix_search)) %>% polloi::make_dygraph(xlab = "Date", ylab = "Searches", title = "Prefix Search API usage by day", legend_name = "Searches") %>% - dyRangeSelector + dyRangeSelector %>% + dyEvent(as.Date("2017-01-01"), "R (reportupdater)", labelLoc = "bottom") }) # Failure plots @@ -284,7 +298,8 @@ dyAxis("y", axisLabelFormatter = "function(x) { return x + '%'; }", valueFormatter = "function(x) { return x + '%'; }") %>% dyRangeSelector(fillColor = "") %>% dyEvent(as.Date("2016-02-01"), "A (format switch)", labelLoc = "bottom") %>% - dyEvent(as.Date("2016-03-16"), "Completion Suggester Deployed", labelLoc = "bottom") + dyEvent(as.Date("2016-03-16"), "Completion Suggester Deployed", labelLoc = "bottom") %>% + dyEvent(as.Date("2017-01-01"), "R (reportupdater)", labelLoc = "bottom") }) output$failure_rate_change_plot <- renderDygraph({ @@ -294,7 +309,8 @@ polloi::make_dygraph(xlab = "Date", ylab = "Change", title = "Zero Results rate change, by day", legend_name = "Change") %>% dyAxis("y", axisLabelFormatter = "function(x) { return x + '%'; }", valueFormatter = "function(x) { return Math.round(x, 3) + '%'; }") %>% dyRangeSelector(fillColor = "", strokeColor = "") %>% - dyEvent(as.Date("2016-03-16"), "Completion Suggester Deployed", labelLoc = "bottom") + dyEvent(as.Date("2016-03-16"), "Completion Suggester Deployed", labelLoc = "bottom") %>% + dyEvent(as.Date("2017-01-01"), "R (reportupdater)", labelLoc = "bottom") }) output$failure_breakdown_plot <- renderDygraph({ @@ -324,7 +340,8 @@ dyRangeSelector(fillColor = "") %>% # Remember to update the tab documentation with details about the annotations! dyEvent(as.Date("2016-02-01"), "A (format switch)", labelLoc = "bottom") %>% - dyEvent(as.Date("2016-03-16"), "Completion Suggester Deployed", labelLoc = "bottom") + dyEvent(as.Date("2016-03-16"), "Completion Suggester Deployed", labelLoc = "bottom") %>% + dyEvent(as.Date("2017-01-01"), "R (reportupdater)", labelLoc = "bottom") }) output$suggestion_dygraph_plot <- renderDygraph({ @@ -335,7 +352,8 @@ dyAxis("y", axisLabelFormatter = "function(x) { return x + '%'; }", valueFormatter = "function(x) { return x + '%'; }") %>% dyRangeSelector(fillColor = "") %>% dyEvent(as.Date("2016-02-01"), "A (format switch)", labelLoc = "bottom") %>% - dyEvent(as.Date("2016-03-16"), "Completion Suggester Deployed", labelLoc = "bottom") + dyEvent(as.Date("2016-03-16"), "Completion Suggester Deployed", labelLoc = "bottom") %>% + dyEvent(as.Date("2017-01-01"), "R (reportupdater)", labelLoc = "bottom") }) output$language_selector_container <- renderUI({ @@ -393,7 +411,8 @@ dyAxis("y", axisLabelFormatter = "function(x) { return x + '%'; }", valueFormatter = "function(x) { return x + '%'; }") %>% dyLegend(show = "always", width = 400, labelsDiv = "failure_langproj_legend") %>% dyAxis("x", axisLabelFormatter = polloi::custom_axis_formatter) %>% - dyRangeSelector(fillColor = "") + dyRangeSelector(fillColor = "") %>% + dyEvent(as.Date("2017-01-01"), "R (reportupdater)", labelLoc = "bottom") }) # Survival @@ -404,7 +423,8 @@ dyAxis("x", ticker = "Dygraph.dateTicker", axisLabelFormatter = polloi::custom_axis_formatter, axisLabelWidth = 100, pixelsPerLabel = 80) %>% dyLegend(labelsDiv = "lethal_dose_plot_legend") %>% - dyRangeSelector(fillColor = "", strokeColor = "") + dyRangeSelector(fillColor = "", strokeColor = "") %>% + dyEvent(as.Date("2017-01-01"), "R (reportupdater)", labelLoc = "bottom") }) ## KPI Summary Boxes @@ -730,7 +750,8 @@ includeZero = TRUE) %>% dyCSS(css = system.file("custom.css", package = "polloi")) %>% dyRangeSelector %>% - dyEvent(as.Date("2016-07-12"), "A (schema switch)", labelLoc = "bottom")) + dyEvent(as.Date("2016-07-12"), "A (schema switch)", labelLoc = "bottom") %>% + dyEvent(as.Date("2017-01-01"), "R (reportupdater)", labelLoc = "bottom")) }) output$kpi_zero_results_series <- renderDygraph({ smooth_level <- input$smoothing_kpi_zero_results @@ -773,7 +794,8 @@ dyCSS(css = system.file("custom.css", package = "polloi")) %>% dyRangeSelector(fillColor = "") %>% dyEvent(as.Date("2016-02-01"), "A (format switch)", labelLoc = "bottom") %>% - dyEvent(as.Date("2016-03-16"), "Completion Suggester Deployed", labelLoc = "bottom")) + dyEvent(as.Date("2016-03-16"), "Completion Suggester Deployed", labelLoc = "bottom") %>% + dyEvent(as.Date("2017-01-01"), "R (reportupdater)", labelLoc = "bottom")) }) output$kpi_api_usage_series <- renderDygraph({ smooth_level <- input$smoothing_kpi_api_usage @@ -810,7 +832,8 @@ logscale = input$kpi_api_usage_series_log_scale ) %>% dyCSS(css = system.file("custom.css", package = "polloi")) %>% - dyRangeSelector) + dyRangeSelector %>% + dyEvent(as.Date("2017-01-01"), "R (reportupdater)", labelLoc = "bottom")) } api_usage_change <- dplyr::mutate( api_usage, @@ -832,7 +855,8 @@ drawPoints = FALSE, pointSize = 3, labelsKMB = TRUE, includeZero = TRUE) %>% dyCSS(css = system.file("custom.css", package = "polloi")) %>% - dyRangeSelector) + dyRangeSelector %>% + dyEvent(as.Date("2017-01-01"), "R (reportupdater)", labelLoc = "bottom")) }) output$kpi_augmented_clickthroughs_series <- renderDygraph({ start_date <- Sys.Date() - switch(input$kpi_summary_date_range_selector, all = NA, daily = 1, weekly = 8, monthly = 31, quarterly = 91) @@ -851,7 +875,8 @@ dyLegend(labelsDiv = "kpi_augmented_clickthroughs_series_legend") %>% dyRangeSelector(fillColor = "", strokeColor = "") %>% dyEvent(as.Date("2016-03-16"), "Completion Suggester Deployed", labelLoc = "bottom") %>% - dyEvent(as.Date("2016-07-12"), "A (schema switch)", labelLoc = "bottom") + dyEvent(as.Date("2016-07-12"), "A (schema switch)", labelLoc = "bottom") %>% + dyEvent(as.Date("2017-01-01"), "R (reportupdater)", labelLoc = "bottom") }) output$monthly_metrics_tbl <- DT::renderDataTable({ diff --git a/tab_documentation/app_events.md b/tab_documentation/app_events.md index 424f7f0..02b87ba 100644 --- a/tab_documentation/app_events.md +++ b/tab_documentation/app_events.md @@ -15,6 +15,8 @@ ------ * There is a spike in events on 2 June 2015 because of a release of the iOS app that added search logging. This has been [confirmed](https://phabricator.wikimedia.org/T102098) by a mobile apps software engineer. +* '__R__': on 2017-01-01 we started calculating all of Discovery's metrics using a new version of [our data retrieval and processing codebase](https://phabricator.wikimedia.org/diffusion/WDGO/) that we migrated to [Wikimedia Analytics](https://www.mediawiki.org/wiki/Analytics)' [Reportupdater infrastructure](https://wikitech.wikimedia.org/wiki/Analytics/Reportupdater). See [T150915](https://phabricator.wikimedia.org/T150915) for more details. + Outages and inaccuracies ------ * Between 5 May and 6 May 2015, approximately 40% of incoming EventLogging data was lost due to a wider EventLogging outage. You can read more about the outage [here](https://wikitech.wikimedia.org/wiki/Incident_documentation/20150506-EventLogging). diff --git a/tab_documentation/app_load.md b/tab_documentation/app_load.md index b39322e..f46fddf 100644 --- a/tab_documentation/app_load.md +++ b/tab_documentation/app_load.md @@ -17,6 +17,7 @@ * Between 5 May and 6 May 2015, approximately 40% of incoming EventLogging data was lost due to a wider EventLogging outage. You can read more about the outage [here](https://wikitech.wikimedia.org/wiki/Incident_documentation/20150506-EventLogging). * A bug in the app implementation led to the clocks being off, hence some claims of -38000 seconds as a completion time. This is now patched. +* '__R__': on 2017-01-01 we started calculating all of Discovery's metrics using a new version of [our data retrieval and processing codebase](https://phabricator.wikimedia.org/diffusion/WDGO/) that we migrated to [Wikimedia Analytics](https://www.mediawiki.org/wiki/Analytics)' [Reportupdater infrastructure](https://wikitech.wikimedia.org/wiki/Analytics/Reportupdater). See [T150915](https://phabricator.wikimedia.org/T150915) for more details. Questions, bug reports, and feature suggestions ------ diff --git a/tab_documentation/click_position.md b/tab_documentation/click_position.md index 17a764c..e15f9cd 100644 --- a/tab_documentation/click_position.md +++ b/tab_documentation/click_position.md @@ -5,6 +5,7 @@ Notes ------ * Currently, we only have data from the latest version of the android app ([Schema:MobileWikiAppSearch, Revision 15729321](https://meta.wikimedia.org/wiki/Schema:MobileWikiAppSearch)). +* '__R__': on 2017-01-01 we started calculating all of Discovery's metrics using a new version of [our data retrieval and processing codebase](https://phabricator.wikimedia.org/diffusion/WDGO/) that we migrated to [Wikimedia Analytics](https://www.mediawiki.org/wiki/Analytics)' [Reportupdater infrastructure](https://wikitech.wikimedia.org/wiki/Analytics/Reportupdater). See [T150915](https://phabricator.wikimedia.org/T150915) for more details. Questions, bug reports, and feature suggestions ------ diff --git a/tab_documentation/desktop_events.md b/tab_documentation/desktop_events.md index 439e6f6..9c07536 100644 --- a/tab_documentation/desktop_events.md +++ b/tab_documentation/desktop_events.md @@ -17,7 +17,8 @@ * Between 2 October 2015 and 28 October 2015 we were not logging any events from the Search schema. There was a change in core that broke the code being inserted into pages. Those pages were cached into varnish so an alternate solution had to be taken. that was delayed because of deployment freezes. The change in core only broke it because the way the code was added from our side was technically wrong, but happened to work anyways. * Between 5 May and 6 May 2015, approximately 40% of incoming EventLogging data was lost due to a wider EventLogging outage. You can read more about the outage [here](https://wikitech.wikimedia.org/wiki/Incident_documentation/20150506-EventLogging). * Data in late September/early October 2015 is unavailable due to another bug in EventLogging as a whole, which impacted data collection. -* __A__: we switched to using data from [Schema:TestSearchSatisfaction2](https://meta.wikimedia.org/wiki/Schema:TestSearchSatisfaction2) instead of [Schema:Search](https://meta.wikimedia.org/wiki/Schema:Search) for Desktop event counts and load times on 12 July 2016. +* '__A__': we switched to using data from [Schema:TestSearchSatisfaction2](https://meta.wikimedia.org/wiki/Schema:TestSearchSatisfaction2) instead of [Schema:Search](https://meta.wikimedia.org/wiki/Schema:Search) for Desktop event counts and load times on 12 July 2016. +* '__R__': on 2017-01-01 we started calculating all of Discovery's metrics using a new version of [our data retrieval and processing codebase](https://phabricator.wikimedia.org/diffusion/WDGO/) that we migrated to [Wikimedia Analytics](https://www.mediawiki.org/wiki/Analytics)' [Reportupdater infrastructure](https://wikitech.wikimedia.org/wiki/Analytics/Reportupdater). See [T150915](https://phabricator.wikimedia.org/T150915) for more details. Questions, bug reports, and feature suggestions ------ diff --git a/tab_documentation/desktop_load.md b/tab_documentation/desktop_load.md index 2a8b122..a005e60 100644 --- a/tab_documentation/desktop_load.md +++ b/tab_documentation/desktop_load.md @@ -17,7 +17,8 @@ * Between 2 October 2015 and 28 October 2015 we were not logging any events from the Search schema. There was a change in core that broke the code being inserted into pages. Those pages were cached into varnish so an alternate solution had to be taken. that was delayed because of deployment freezes. The change in core only broke it because the way the code was added from our side was technically wrong, but happened to work anyways. * Between 5 May and 6 May 2015, approximately 40% of incoming EventLogging data was lost due to a wider EventLogging outage. You can read more about the outage [here](https://wikitech.wikimedia.org/wiki/Incident_documentation/20150506-EventLogging). * Data in late September/early October 2015 is unavailable due to another bug in EventLogging as a whole, which impacted data collection. -* __A__: we switched to using data from [Schema:TestSearchSatisfaction2](https://meta.wikimedia.org/wiki/Schema:TestSearchSatisfaction2) instead of [Schema:Search](https://meta.wikimedia.org/wiki/Schema:Search) for Desktop event counts and load times on 12 July 2016. +* '__A__': we switched to using data from [Schema:TestSearchSatisfaction2](https://meta.wikimedia.org/wiki/Schema:TestSearchSatisfaction2) instead of [Schema:Search](https://meta.wikimedia.org/wiki/Schema:Search) for Desktop event counts and load times on 12 July 2016. +* '__R__': on 2017-01-01 we started calculating all of Discovery's metrics using a new version of [our data retrieval and processing codebase](https://phabricator.wikimedia.org/diffusion/WDGO/) that we migrated to [Wikimedia Analytics](https://www.mediawiki.org/wiki/Analytics)' [Reportupdater infrastructure](https://wikitech.wikimedia.org/wiki/Analytics/Reportupdater). See [T150915](https://phabricator.wikimedia.org/T150915) for more details. Questions, bug reports, and feature suggestions ------ diff --git a/tab_documentation/failure_breakdown.md b/tab_documentation/failure_breakdown.md index 1a69092..831d309 100644 --- a/tab_documentation/failure_breakdown.md +++ b/tab_documentation/failure_breakdown.md @@ -6,9 +6,10 @@ Notes ------ -* __Anotation "A"__: On 13 April 2016 we switched to a new data format for zero results rate (see [T132503](https://phabricator.wikimedia.org/T132503)) wherein we stopped lumping in different query types into just two categories ("Full-Text Search" and "Prefix Search"). The ZRR data were backfilled from 1 February 2016 under the new format which breaks down ZRR into the individual query types. We also began filtering out irrelevant query types (see [T131196#2200560](https://phabricator.wikimedia.org/T131196#2200560)) and requests with unknown number of hits ("-1" in the database). +* '__A__': On 13 April 2016 we switched to a new data format for zero results rate (see [T132503](https://phabricator.wikimedia.org/T132503)) wherein we stopped lumping in different query types into just two categories ("Full-Text Search" and "Prefix Search"). The ZRR data were backfilled from 1 February 2016 under the new format which breaks down ZRR into the individual query types. We also began filtering out irrelevant query types (see [T131196#2200560](https://phabricator.wikimedia.org/T131196#2200560)) and requests with unknown number of hits ("-1" in the database). * On 15 July 2015 we updated our heuristics to avoid counting maintenance tasks as search requests. * On 15 January 2016 there was an [issue](https://phabricator.wikimedia.org/T123541) with Avro serialization that prevented data from entering the Hadoop cluster. A [patch](https://gerrit.wikimedia.org/r/#/c/264989/) was deployed on 19 January 2016. As a result, there are no recorded zero results rates for 01/15-01/19. The values you may see on those dates are estimates computed with [statistical models](https://github.com/bearloga/branch/blob/master/zero%20results%20rate%20estimation/report.pdf). +* '__R__': on 2017-01-01 we started calculating all of Discovery's metrics using a new version of [our data retrieval and processing codebase](https://phabricator.wikimedia.org/diffusion/WDGO/) that we migrated to [Wikimedia Analytics](https://www.mediawiki.org/wiki/Analytics)' [Reportupdater infrastructure](https://wikitech.wikimedia.org/wiki/Analytics/Reportupdater). See [T150915](https://phabricator.wikimedia.org/T150915) for more details. Questions, bug reports, and feature suggestions ------ diff --git a/tab_documentation/failure_langproj.md b/tab_documentation/failure_langproj.md index 723eaa8..42de070 100644 --- a/tab_documentation/failure_langproj.md +++ b/tab_documentation/failure_langproj.md @@ -16,6 +16,7 @@ Outages and inaccuracies ------ * On 15 January 2016 there was an [issue](https://phabricator.wikimedia.org/T123541) with Avro serialization that prevented data from entering the Hadoop cluster. A [patch](https://gerrit.wikimedia.org/r/#/c/264989/) was deployed on 19 January 2016. As a result, there are no recorded zero results rates for 01/15-01/19. The values you may see on those dates are estimates computed with [statistical models](https://github.com/bearloga/branch/blob/master/zero%20results%20rate%20estimation/report.pdf). +* '__R__': on 2017-01-01 we started calculating all of Discovery's metrics using a new version of [our data retrieval and processing codebase](https://phabricator.wikimedia.org/diffusion/WDGO/) that we migrated to [Wikimedia Analytics](https://www.mediawiki.org/wiki/Analytics)' [Reportupdater infrastructure](https://wikitech.wikimedia.org/wiki/Analytics/Reportupdater). See [T150915](https://phabricator.wikimedia.org/T150915) for more details. Questions, bug reports, and feature suggestions ------ diff --git a/tab_documentation/failure_rate.md b/tab_documentation/failure_rate.md index 77d1557..5494cae 100644 --- a/tab_documentation/failure_rate.md +++ b/tab_documentation/failure_rate.md @@ -10,9 +10,10 @@ Notes ------ -* __Anotation "A"__: On 13 April 2016 we switched to a new data format for zero results rate (see [T132503](https://phabricator.wikimedia.org/T132503)) wherein we stopped lumping in different query types into just two categories ("Full-Text Search" and "Prefix Search"). The ZRR data were backfilled from 1 February 2016 under the new format which breaks down ZRR into the individual query types. We also began filtering out irrelevant query types (see [T131196#2200560](https://phabricator.wikimedia.org/T131196#2200560)) and requests with unknown number of hits ("-1" in the database). +* '__Anotation A__': On 13 April 2016 we switched to a new data format for zero results rate (see [T132503](https://phabricator.wikimedia.org/T132503)) wherein we stopped lumping in different query types into just two categories ("Full-Text Search" and "Prefix Search"). The ZRR data were backfilled from 1 February 2016 under the new format which breaks down ZRR into the individual query types. We also began filtering out irrelevant query types (see [T131196#2200560](https://phabricator.wikimedia.org/T131196#2200560)) and requests with unknown number of hits ("-1" in the database). * On 15 July 2015 we updated our heuristics to avoid counting maintenance tasks as search requests. The historic data on the dashboards is being backfilled to reflect this - until it's done, the dashboards may look somewhat strange. * On 15 January 2016 there was an [issue](https://phabricator.wikimedia.org/T123541) with Avro serialization that prevented data from entering the Hadoop cluster. A [patch](https://gerrit.wikimedia.org/r/#/c/264989/) was deployed on 19 January 2016. As a result, there are no recorded zero results rates for 01/15-01/19. The values you may see on those dates are estimates computed with [statistical models](https://github.com/bearloga/branch/blob/master/zero%20results%20rate%20estimation/report.pdf). +* '__R__': on 2017-01-01 we started calculating all of Discovery's metrics using a new version of [our data retrieval and processing codebase](https://phabricator.wikimedia.org/diffusion/WDGO/) that we migrated to [Wikimedia Analytics](https://www.mediawiki.org/wiki/Analytics)' [Reportupdater infrastructure](https://wikitech.wikimedia.org/wiki/Analytics/Reportupdater). See [T150915](https://phabricator.wikimedia.org/T150915) for more details. Questions, bug reports, and feature suggestions ------ diff --git a/tab_documentation/failure_suggests.md b/tab_documentation/failure_suggests.md index a0a9363..9870696 100644 --- a/tab_documentation/failure_suggests.md +++ b/tab_documentation/failure_suggests.md @@ -8,7 +8,8 @@ Notes ------ -* __Anotation "A"__: On 13 April 2016 we switched to a new data format for zero results rate (see [T132503](https://phabricator.wikimedia.org/T132503)) wherein we stopped lumping in different query types into just two categories ("Full-Text Search" and "Prefix Search"). The ZRR data were backfilled from 1 February 2016 under the new format which breaks down ZRR into the individual query types. We also began filtering out irrelevant query types (see [T131196#2200560](https://phabricator.wikimedia.org/T131196#2200560)) and requests with unknown number of hits ("-1" in the database). +* '__A__': On 13 April 2016 we switched to a new data format for zero results rate (see [T132503](https://phabricator.wikimedia.org/T132503)) wherein we stopped lumping in different query types into just two categories ("Full-Text Search" and "Prefix Search"). The ZRR data were backfilled from 1 February 2016 under the new format which breaks down ZRR into the individual query types. We also began filtering out irrelevant query types (see [T131196#2200560](https://phabricator.wikimedia.org/T131196#2200560)) and requests with unknown number of hits ("-1" in the database). +* '__R__': on 2017-01-01 we started calculating all of Discovery's metrics using a new version of [our data retrieval and processing codebase](https://phabricator.wikimedia.org/diffusion/WDGO/) that we migrated to [Wikimedia Analytics](https://www.mediawiki.org/wiki/Analytics)' [Reportupdater infrastructure](https://wikitech.wikimedia.org/wiki/Analytics/Reportupdater). See [T150915](https://phabricator.wikimedia.org/T150915) for more details. Questions, bug reports, and feature suggestions ------ diff --git a/tab_documentation/fulltext_basic.md b/tab_documentation/fulltext_basic.md index f120b56..f536caa 100644 --- a/tab_documentation/fulltext_basic.md +++ b/tab_documentation/fulltext_basic.md @@ -12,7 +12,7 @@ Outages and inaccuracies ------ -* None so far! +* '__R__': on 2017-01-01 we started calculating all of Discovery's metrics using a new version of [our data retrieval and processing codebase](https://phabricator.wikimedia.org/diffusion/WDGO/) that we migrated to [Wikimedia Analytics](https://www.mediawiki.org/wiki/Analytics)' [Reportupdater infrastructure](https://wikitech.wikimedia.org/wiki/Analytics/Reportupdater). See [T150915](https://phabricator.wikimedia.org/T150915) for more details. Furthermore, we switched to an updated UDF for counting API calls -- the previous version was undercounting full-text and geo search API calls (see [Gerrit change 315503](https://gerrit.wikimedia.org/r/#/c/315503/) for more details). Questions, bug reports, and feature suggestions ------ diff --git a/tab_documentation/geo_basic.md b/tab_documentation/geo_basic.md index 5cefe85..188a450 100644 --- a/tab_documentation/geo_basic.md +++ b/tab_documentation/geo_basic.md @@ -1,21 +1,17 @@ Geographic Search API usage ======= -Geo Search, or Geographic Search, refers to the ability to search for pages that are "near" a particular set of geographic coordinates, -in the sense of being about subjects that have physical locations we track. - -(places. Not people.) +Geo Search, or Geographic Search, refers to the ability to search for pages that are "near" a particular set of geographic coordinates, in the sense of being about subjects that have physical locations we track. (places. Not people.) General trends ------ -It's early days, but the tiny adoption of Geographic Search, particularly compared to Cirrus or OpenSearch, suggests this isn't a tremendously useful feature to API users. We need to dig in and refine the heuristics for identifying requests as different types, -however. +It's early days, but the tiny adoption of Geographic Search, particularly compared to Cirrus or OpenSearch, suggests this isn't a tremendously useful feature to API users. We need to dig in and refine the heuristics for identifying requests as different types, however. Outages and inaccuracies ------ - * None so far! +* '__R__': on 2017-01-01 we started calculating all of Discovery's metrics using a new version of [our data retrieval and processing codebase](https://phabricator.wikimedia.org/diffusion/WDGO/) that we migrated to [Wikimedia Analytics](https://www.mediawiki.org/wiki/Analytics)' [Reportupdater infrastructure](https://wikitech.wikimedia.org/wiki/Analytics/Reportupdater). See [T150915](https://phabricator.wikimedia.org/T150915) for more details. Furthermore, we switched to an updated UDF for counting API calls -- the previous version was undercounting full-text and geo search API calls (see [Gerrit change 315503](https://gerrit.wikimedia.org/r/#/c/315503/) for more details). Questions, bug reports, and feature suggestions ------ diff --git a/tab_documentation/invoke_source.md b/tab_documentation/invoke_source.md index 184fdfd..59ef97f 100644 --- a/tab_documentation/invoke_source.md +++ b/tab_documentation/invoke_source.md @@ -5,6 +5,7 @@ Notes ------ * Currently, we only have data from the latest version of the android app ([Schema:MobileWikiAppSearch, Revision 15729321](https://meta.wikimedia.org/wiki/Schema:MobileWikiAppSearch)). +* '__R__': on 2017-01-01 we started calculating all of Discovery's metrics using a new version of [our data retrieval and processing codebase](https://phabricator.wikimedia.org/diffusion/WDGO/) that we migrated to [Wikimedia Analytics](https://www.mediawiki.org/wiki/Analytics)' [Reportupdater infrastructure](https://wikitech.wikimedia.org/wiki/Analytics/Reportupdater). See [T150915](https://phabricator.wikimedia.org/T150915) for more details. Questions, bug reports, and feature suggestions ------ diff --git a/tab_documentation/kpi_api_usage.md b/tab_documentation/kpi_api_usage.md index c9971f6..3044265 100644 --- a/tab_documentation/kpi_api_usage.md +++ b/tab_documentation/kpi_api_usage.md @@ -6,7 +6,7 @@ Outages and inaccuracies ------ -* None so far! +* '__R__': on 2017-01-01 we started calculating all of Discovery's metrics using a new version of [our data retrieval and processing codebase](https://phabricator.wikimedia.org/diffusion/WDGO/) that we migrated to [Wikimedia Analytics](https://www.mediawiki.org/wiki/Analytics)' [Reportupdater infrastructure](https://wikitech.wikimedia.org/wiki/Analytics/Reportupdater). See [T150915](https://phabricator.wikimedia.org/T150915) for more details. Furthermore, we switched to an updated UDF for counting API calls -- the previous version was undercounting full-text and geo search API calls (see [Gerrit change 315503](https://gerrit.wikimedia.org/r/#/c/315503/) for more details). Questions, bug reports, and feature suggestions ------ diff --git a/tab_documentation/kpi_augmented_clickthroughs.md b/tab_documentation/kpi_augmented_clickthroughs.md index c261e31..fd07351 100644 --- a/tab_documentation/kpi_augmented_clickthroughs.md +++ b/tab_documentation/kpi_augmented_clickthroughs.md @@ -7,7 +7,8 @@ Notes ------ -* __A__: we switched to using data from [Schema:TestSearchSatisfaction2](https://meta.wikimedia.org/wiki/Schema:TestSearchSatisfaction2) instead of [Schema:Search](https://meta.wikimedia.org/wiki/Schema:Search) for Desktop event counts and load times on 12 July 2016. +* '__A__': we switched to using data from [Schema:TestSearchSatisfaction2](https://meta.wikimedia.org/wiki/Schema:TestSearchSatisfaction2) instead of [Schema:Search](https://meta.wikimedia.org/wiki/Schema:Search) for Desktop event counts and load times on 12 July 2016. +* '__R__': on 2017-01-01 we started calculating all of Discovery's metrics using a new version of [our data retrieval and processing codebase](https://phabricator.wikimedia.org/diffusion/WDGO/) that we migrated to [Wikimedia Analytics](https://www.mediawiki.org/wiki/Analytics)' [Reportupdater infrastructure](https://wikitech.wikimedia.org/wiki/Analytics/Reportupdater). See [T150915](https://phabricator.wikimedia.org/T150915) for more details. Questions, bug reports, and feature suggestions ------ diff --git a/tab_documentation/kpi_load_time.md b/tab_documentation/kpi_load_time.md index edc090b..6b53944 100644 --- a/tab_documentation/kpi_load_time.md +++ b/tab_documentation/kpi_load_time.md @@ -5,7 +5,8 @@ Notes ------ -* __A__: we switched to using data from [Schema:TestSearchSatisfaction2](https://meta.wikimedia.org/wiki/Schema:TestSearchSatisfaction2) instead of [Schema:Search](https://meta.wikimedia.org/wiki/Schema:Search) for Desktop event counts and load times on 12 July 2016. +* '__A__': we switched to using data from [Schema:TestSearchSatisfaction2](https://meta.wikimedia.org/wiki/Schema:TestSearchSatisfaction2) instead of [Schema:Search](https://meta.wikimedia.org/wiki/Schema:Search) for Desktop event counts and load times on 12 July 2016. +* '__R__': on 2017-01-01 we started calculating all of Discovery's metrics using a new version of [our data retrieval and processing codebase](https://phabricator.wikimedia.org/diffusion/WDGO/) that we migrated to [Wikimedia Analytics](https://www.mediawiki.org/wiki/Analytics)' [Reportupdater infrastructure](https://wikitech.wikimedia.org/wiki/Analytics/Reportupdater). See [T150915](https://phabricator.wikimedia.org/T150915) for more details. Questions, bug reports, and feature suggestions ------ diff --git a/tab_documentation/kpi_zero_results.md b/tab_documentation/kpi_zero_results.md index e72326a..81e3e71 100644 --- a/tab_documentation/kpi_zero_results.md +++ b/tab_documentation/kpi_zero_results.md @@ -8,6 +8,7 @@ * __Anotation "A"__: On 13 April 2016 we switched to a new data format for zero results rate (see [T132503](https://phabricator.wikimedia.org/T132503)) wherein we stopped lumping in different query types into just two categories ("Full-Text Search" and "Prefix Search"). The ZRR data were backfilled from 1 February 2016 under the new format which breaks down ZRR into the individual query types. We also began filtering out irrelevant query types (see [T131196#2200560](https://phabricator.wikimedia.org/T131196#2200560)) and requests with unknown number of hits ("-1" in the database). * On 15 January 2016 there was an [issue](https://phabricator.wikimedia.org/T123541) with Avro serialization that prevented data from entering the Hadoop cluster. A [patch](https://gerrit.wikimedia.org/r/#/c/264989/) was deployed on 19 January 2016. As a result, there are no recorded zero results rates for 01/15-01/19. The values you may see on those dates are estimates computed with [statistical models](https://github.com/bearloga/branch/blob/master/zero%20results%20rate%20estimation/report.pdf). +* '__R__': on 2017-01-01 we started calculating all of Discovery's metrics using a new version of [our data retrieval and processing codebase](https://phabricator.wikimedia.org/diffusion/WDGO/) that we migrated to [Wikimedia Analytics](https://www.mediawiki.org/wiki/Analytics)' [Reportupdater infrastructure](https://wikitech.wikimedia.org/wiki/Analytics/Reportupdater). See [T150915](https://phabricator.wikimedia.org/T150915) for more details. Questions, bug reports, and feature suggestions ------ diff --git a/tab_documentation/language_basic.md b/tab_documentation/language_basic.md index 593eda1..bf56269 100644 --- a/tab_documentation/language_basic.md +++ b/tab_documentation/language_basic.md @@ -11,7 +11,7 @@ Outages and inaccuracies ------ - * None so far! +* '__R__': on 2017-01-01 we started calculating all of Discovery's metrics using a new version of [our data retrieval and processing codebase](https://phabricator.wikimedia.org/diffusion/WDGO/) that we migrated to [Wikimedia Analytics](https://www.mediawiki.org/wiki/Analytics)' [Reportupdater infrastructure](https://wikitech.wikimedia.org/wiki/Analytics/Reportupdater). See [T150915](https://phabricator.wikimedia.org/T150915) for more details. Questions, bug reports, and feature suggestions ------ diff --git a/tab_documentation/mobile_events.md b/tab_documentation/mobile_events.md index 8908774..5bd522a 100644 --- a/tab_documentation/mobile_events.md +++ b/tab_documentation/mobile_events.md @@ -11,14 +11,16 @@ General trends ------ -It's hard to tell because we have too little data and there's clearly something screwy in the data we're provided with. +It's hard to tell because we have too little data and there's clearly something screwy in the data we're provided with. Outages and inaccuracies ------ + There are occasionally going to be outages that will affect the accuracy of data. To make it easier to rely on the data (or not!) they are listed here, from most- to least-recent. * Between 5 May and 6 May 2015, approximately 40% of incoming EventLogging data was lost due to a wider EventLogging outage. You can read more about the outage [here](https://wikitech.wikimedia.org/wiki/Incident_documentation/20150506-EventLogging). +* '__R__': on 2017-01-01 we started calculating all of Discovery's metrics using a new version of [our data retrieval and processing codebase](https://phabricator.wikimedia.org/diffusion/WDGO/) that we migrated to [Wikimedia Analytics](https://www.mediawiki.org/wiki/Analytics)' [Reportupdater infrastructure](https://wikitech.wikimedia.org/wiki/Analytics/Reportupdater). See [T150915](https://phabricator.wikimedia.org/T150915) for more details. Questions, bug reports, and feature suggestions ------ diff --git a/tab_documentation/mobile_load.md b/tab_documentation/mobile_load.md index 7efd7bc..c51ec4f 100644 --- a/tab_documentation/mobile_load.md +++ b/tab_documentation/mobile_load.md @@ -10,13 +10,13 @@ ------ It's hard to tell because we have too little data and there's clearly something screwy in the data we're provided with. Like: seriously screwy. - Outages and inaccuracies ------ There are occasionally going to be outages that will affect the accuracy of data. To make it easier to rely on the data (or not!) they are listed here, from most- to least-recent. * Between 5 May and 6 May 2015, approximately 40% of incoming EventLogging data was lost due to a wider EventLogging outage. You can read more about the outage [here](https://wikitech.wikimedia.org/wiki/Incident_documentation/20150506-EventLogging). * There is another one and god knows where it's from. Just look at that graph. +* '__R__': on 2017-01-01 we started calculating all of Discovery's metrics using a new version of [our data retrieval and processing codebase](https://phabricator.wikimedia.org/diffusion/WDGO/) that we migrated to [Wikimedia Analytics](https://www.mediawiki.org/wiki/Analytics)' [Reportupdater infrastructure](https://wikitech.wikimedia.org/wiki/Analytics/Reportupdater). See [T150915](https://phabricator.wikimedia.org/T150915) for more details. Questions, bug reports, and feature suggestions ------ diff --git a/tab_documentation/open_basic.md b/tab_documentation/open_basic.md index 3c64c11..e017330 100644 --- a/tab_documentation/open_basic.md +++ b/tab_documentation/open_basic.md @@ -14,7 +14,7 @@ Outages and inaccuracies ------ - * None so far! +* '__R__': on 2017-01-01 we started calculating all of Discovery's metrics using a new version of [our data retrieval and processing codebase](https://phabricator.wikimedia.org/diffusion/WDGO/) that we migrated to [Wikimedia Analytics](https://www.mediawiki.org/wiki/Analytics)' [Reportupdater infrastructure](https://wikitech.wikimedia.org/wiki/Analytics/Reportupdater). See [T150915](https://phabricator.wikimedia.org/T150915) for more details. Questions, bug reports, and feature suggestions ------ diff --git a/tab_documentation/paulscore_approx.html b/tab_documentation/paulscore_approx.html index 0661123..0828508 100644 --- a/tab_documentation/paulscore_approx.html +++ b/tab_documentation/paulscore_approx.html @@ -1,29 +1,35 @@ <h1>PaulScore Approximations</h1> <div> -<div style = "float: left; width: 47%; padding: 0 3% 0 1%;"> -<p>"PaulScore" is the name we've given to a metric proposed by Paul Nelson in a talk he gave at <a href = "https://www.youtube.com/watch?v=YJ_amC9gZmk&t=16m38s" title = "Paul Nelson's talk at Elasticon">Elasticon</a>. We use PaulScore to evaluate the quality of results provided by CirrusSearch or proposed modifications to CirrusSearch, based on historical click data. A big advantage of the PaulScore is that it relies on user click history to award points, so it is easy to compute.</p> +<div style="float: left; width: 47%; padding: 0 3% 0 1%;"> +<p>"PaulScore" is the name we've given to a metric proposed by Paul Nelson in a talk he gave at <a href="https://www.youtube.com/watch?v=YJ_amC9gZmk&t=16m38s" title="Paul Nelson's talk at Elasticon">Elasticon</a>. We use PaulScore to evaluate the quality of results provided by CirrusSearch or proposed modifications to CirrusSearch, based on historical click data. A big advantage of the PaulScore is that it relies on user click history to award points, so it is easy to compute.</p> <p>This dashboard shows the PaulScore approximation for 3 values of $F$: 0.1, 0.5, and 0.9. The maximum score possible for each value of $F$ is $1/(1-F)$, so the dashboard has the option of looking at relative PaulScores, which is the computed value divided by maximum possible value for given $F$.</p> <p>For auto-completion suggestions, we expect a much lower score since most queries get no clicks -- i.e., while typing many results are shown and ignored -- and most users will only click on one results, whereas full-text searchers can more easily go back to the results page or open multiple results in other windows.</p> -<p>For more details, please see <a href = "https://www.mediawiki.org/wiki/Wikimedia_Discovery/Search/Glossary#PaulScore" title = "Definition of PaulScore">Discovery's Search glossary</a>.</p> +<p>For more details, please see <a href="https://www.mediawiki.org/wiki/Wikimedia_Discovery/Search/Glossary#PaulScore" title="Definition of PaulScore">Discovery's Search glossary</a>.</p> </div> -<div style = "float: left; width: 46%; padding: 0 1% 0 3%;"> +<div style="float: left; width: 46%; padding: 0 1% 0 3%;"> <p>PaulScore is computed via the following steps:</p> <ol> <li>Pick scoring factor $0 < F < 1$.</li> -<li>For $i$-th search session $S_i$ $(i = 1, \ldots, n)$ containing $m$ queries $Q_1, \ldots, Q_m$ and search result sets $\mathbf{R}_1, \ldots, \mathbf{R}_m$: <ol> - <li>For each $j$-th search query $Q_j$ with result set $\mathbf{R}_j$, let $\nu_j$ be the query score: $$\nu_j = \sum_{k~\in~\{\text{0-based positions of clicked results in}~\mathbf{R}_j\}} F^k.$$</li> - <li>Let user's average query score $\bar{\nu}_{(i)}$ be $$\bar{\nu}_{(i)} = \frac{1}{m} \sum_{j = 1}^m \nu_j.$$</li> +<li>For $i$-th search session $S_i$ $(i=1, \ldots, n)$ containing $m$ queries $Q_1, \ldots, Q_m$ and search result sets $\mathbf{R}_1, \ldots, \mathbf{R}_m$: <ol> + <li>For each $j$-th search query $Q_j$ with result set $\mathbf{R}_j$, let $\nu_j$ be the query score: $$\nu_j=\sum_{k~\in~\{\text{0-based positions of clicked results in}~\mathbf{R}_j\}} F^k.$$</li> + <li>Let user's average query score $\bar{\nu}_{(i)}$ be $$\bar{\nu}_{(i)}=\frac{1}{m} \sum_{j=1}^m \nu_j.$$</li> </ol></li> -<li>Then the PaulScore is the average of all users' average query scores: $$\text{PaulScore}~=~\frac{1}{n} \sum_{i = 1}^n \bar{\nu}_{(i)}.$$</li> +<li>Then the PaulScore is the average of all users' average query scores: $$\text{PaulScore}~=~\frac{1}{n} \sum_{i=1}^n \bar{\nu}_{(i)}.$$</li> </ol> </div> </div> -<div style = "clear: both;"></div> +<div style="clear: both;"></div> + +<h2>Outages and inaccuracies</h2> + +<ul> + <li>'<strong>R</strong>': on 2017-01-01 we started calculating all of Discovery's metrics using a new version of <a href="https://phabricator.wikimedia.org/diffusion/WDGO/">our data retrieval and processing codebase</a> that we migrated to <a href="https://www.mediawiki.org/wiki/Analytics">Wikimedia Analytics</a>' <a href="https://wikitech.wikimedia.org/wiki/Analytics/Reportupdater">Reportupdater infrastructure</a>. See <a href="https://phabricator.wikimedia.org/T150915">T150915</a> for more details.</li> +</ul> <h2>Questions, bug reports, and feature suggestions</h2> -<p>For technical, non-bug questions, <a href = "mailto:mpo...@wikimedia.org?subject=Dashboard%20Question">email Mikhail</a>. If you experience a bug or notice something wrong or have a suggestion, <a href = "https://phabricator.wikimedia.org/maniphest/task/create/?projects=Discovery">open a ticket in Phabricator</a> in the Discovery board or <a href = "mailto:d...@wikimedia.org?subject=Dashboard%20Question">email Deb</a>.</p> +<p>For technical, non-bug questions, <a href="mailto:mpo...@wikimedia.org?subject=Dashboard%20Question">email Mikhail</a>. If you experience a bug or notice something wrong or have a suggestion, <a href="https://phabricator.wikimedia.org/maniphest/task/create/?projects=Discovery">open a ticket in Phabricator</a> in the Discovery board or <a href="mailto:d...@wikimedia.org?subject=Dashboard%20Question">email Deb</a>.</p> <hr style="border-color: gray;"> <p style="font-size: small; color: gray;"> diff --git a/tab_documentation/prefix_basic.md b/tab_documentation/prefix_basic.md index b40987d..148ed1c 100644 --- a/tab_documentation/prefix_basic.md +++ b/tab_documentation/prefix_basic.md @@ -11,6 +11,7 @@ Notes ------ * After learning of a change to the search API call, we patched the Analytics Hive UDF refinery (see [287264](https://gerrit.wikimedia.org/r/#/c/287264/)) to check for generator=prefixsearch as well, not just list=prefixsearch. The data was backfilled from 20 March 2016 using the updated Prefix API detection. +* '__R__': on 2017-01-01 we started calculating all of Discovery's metrics using a new version of [our data retrieval and processing codebase](https://phabricator.wikimedia.org/diffusion/WDGO/) that we migrated to [Wikimedia Analytics](https://www.mediawiki.org/wiki/Analytics)' [Reportupdater infrastructure](https://wikitech.wikimedia.org/wiki/Analytics/Reportupdater). See [T150915](https://phabricator.wikimedia.org/T150915) for more details. Questions, bug reports, and feature suggestions ------ diff --git a/tab_documentation/survival.md b/tab_documentation/survival.md index e1004e4..6adea35 100644 --- a/tab_documentation/survival.md +++ b/tab_documentation/survival.md @@ -9,8 +9,7 @@ Outages and inaccuracies ------ - -* None so far! +* '__R__': on 2017-01-01 we started calculating all of Discovery's metrics using a new version of [our data retrieval and processing codebase](https://phabricator.wikimedia.org/diffusion/WDGO/) that we migrated to [Wikimedia Analytics](https://www.mediawiki.org/wiki/Analytics)' [Reportupdater infrastructure](https://wikitech.wikimedia.org/wiki/Analytics/Reportupdater). See [T150915](https://phabricator.wikimedia.org/T150915) for more details. Questions, bug reports, and feature suggestions ------ -- To view, visit https://gerrit.wikimedia.org/r/341742 To unsubscribe, visit https://gerrit.wikimedia.org/r/settings Gerrit-MessageType: newchange Gerrit-Change-Id: Ie650b1eb0f5c9cc40e43a316b71f44e0b8b8cab7 Gerrit-PatchSet: 1 Gerrit-Project: wikimedia/discovery/rainbow Gerrit-Branch: master Gerrit-Owner: Bearloga <mpo...@wikimedia.org> _______________________________________________ MediaWiki-commits mailing list MediaWiki-commits@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/mediawiki-commits