From fc479c6b49297fc17b69fce3170c877156d4824c Mon Sep 17 00:00:00 2001 From: Chris Larsen Date: Mon, 5 Aug 2013 14:50:33 -0400 Subject: [PATCH] Update NEWS for v2.0.0 RC1 --- NEWS | 30 +++++++++++++++++++++++++++++- 1 file changed, 29 insertions(+), 1 deletion(-) diff --git a/NEWS b/NEWS index 17c9002829..0ced395787 100644 --- a/NEWS +++ b/NEWS @@ -1,6 +1,6 @@ OpenTSDB - User visible changes. -* Version 2.0.0 (2013-07-?) +* Version 2.0.0 RC1 (2013-08-05) Noteworthy changes: - Configuration can be provided in a properties file @@ -13,7 +13,35 @@ Noteworthy changes: - New formalized HTTP API, deprecates many of the old HTTP API calls but it is still backwards compatible - New store data points over HTTP via JSON + - New annotations for recording meta data along with data points in a time + series + - New global annotations to record meta data at a specific time but not + associated with a specific series + - New meta data for metrics, tag names, tag values and time series - New optional chunked encoding support for HTTP requests, configurable + - Millisecond resolution support for writing data points and annotations + - Variable length encoding for integer data points, any where from 1 to 8 + bytes instead of using 8 bytes for every point + - CORs support for the HTTP API + - New data injest plugin framework allowing support for different data + formats + - Search plugin framework to push meta data to a search engine for access + - Real-Time publisher framework to publish data points to a secondary system + as soon as they are received at the TSD + - New aggregation functions with alternatives to interpolation including: + - zero if missing sum: returns a 0 if a data point doesn't exist + - max if missing min: returns the maximum value if a data point is missing + - min if missing max: returns the minimum value if a data point is missing + - New TSUID tracking that writes a counter to a new table to track the + different time series stored and how many data points were written + - New meta data trees that allow for flattening time series into a + heirarchical representation similar to a file system + - New meta and tree synchronization CLI tools + - New statistics showing the number of UIDs used and available for each type + - New statistics for the number of current connections to the TSD + - New options for working with rate calculations on counters to rollover + or reset on anomallys + - New Debian package compilable from the source * Version 1.1.1 (2013-??-??) [???????]