Open
Conversation
#1458) * For branch next, add an expression function named FirstDifference, which calculates the first difference of a time series. I noticed there is MovingAverage calculation, so I thought maybe I can enrich the mathematics functions into that. * add some unit tests for FirstDifference
Signed-off-by: Chris Larsen <clarsen@yahoo-inc.com>
Signed-off-by: Chris Larsen <clarsen@yahoo-inc.com>
Co-authored-by: Ion DULGHERU <ion.dulgheru@gmail.com>
…#2034) * Jackson has a serious security problem in 2.9.5, which will cause RCE FasterXML/jackson-databind#2295 * Jackson has a serious security problem in 2.9.5, which will cause RCE FasterXML/jackson-databind#2295 Co-authored-by: chi-chi weng <949409306@qq.com>
Co-authored-by: Itamar Turner-Trauring <itamar@itamarst.org>
Co-authored-by: Itamar Turner-Trauring <itamar@itamarst.org>
* Jackson has a serious security problem in 2.9.5, which will cause RCE FasterXML/jackson-databind#2295 * Jackson has a serious security problem in 2.9.5, which will cause RCE FasterXML/jackson-databind#2295 Co-authored-by: chi-chi weng <949409306@qq.com>
Signed-off-by: Chris Larsen <clarsen@yahoo-inc.com>
Enhanced check_tsd script evaluates each individual metric group separately when given a filter
* Fix SaltScanner race condition on spans maps * Fix 1.6 compatibility
UTs were not adjusted.
…x filter and properly ignore rows that don't match the explicit filter. Also sort the fuzzy filter list in ascending order and implement a static comparator instead of instantiating one on each call.
Fixes a concurrency bug where scanners report their results into a map and would overwrite each other's results Resolves: #1753
* renamed instancename of logger The previous name was copied from another script, cosmetic change only * Change behaviour of --ignore-recent option Previous option would fetch data from opentsdb from --duration seconds ago to time.now(), and then try to remove timestamps that was inside the --ignore-recent seconds ago, however the logic was flawed and it actually only included these seconds. Furthermore opentsdb supports setting an "end" parameter, so we use this to only get the data we want. for example -d 180 -I 80, would render a query parameter that looks like `?start=180s-ago&end=80s-ago`. Keeps it simple. Also added debuglogging to output the actual query sent to OpenTSDB if --debug option is enabled. * fixed logic of --percent-over parameter Previous behaviour didn't work due to wrong logic, would set "crit" or "warn" to True regardless. This change fixes that. * better output from logging Add logmessages to be consistent across alerting-scenarios, and changed format of some floats. Fixed a log messaged that displayed "crit" value where it should have been "warn" value. * Fixed bug in logic that parses results Removed an if statement that `continue`:ed the for-loop if a result was neither a `crit` or `warn` already, however this check also made the logic skip the test to see if no values were returned by opentsdb and -A flag was specified to alert in such scenarios. * changed check for timestamps type Previous behaviour was to check if a timestamp could be cast as a float, which is a bit weird, because opentsdb will return integers. I do doubt that opentsdb would return a timestamp that is not an integer to begin with, so i suspect this check is redundant, but leaving it in for now regardless, as per discussion in PR.
…1895) Co-authored-by: Tony Di Nucci <tony.dinucci@skyscanner.net>
* Add an SLA config flag for rollup intervals Adds a configuration option for rollup intervals to specify their maximum acceptable delay. Queries that cover a time between now and that maximum delay will need to query other tables for that time interval. * Add global config flag to enable splitting queries Adds a global config flag to enable splitting queries that would hit the rollup table, but the rollup table has a delay SLA configured. In that case, this feature allows splitting a query into to; one that gets the data from the rollups table until the time where it's guaranteed to be available, and the rest from the raw table. * Add a new SplitRollupQuery Adds a SplitRollupQuery class that suports splitting a rollup query into two separate queries. This is useful for when a rollup table is filled by e.g. a batch job that processes the data from the previous day on a daily basis. Rollup data for yesterday will then only be available some time today. This delay SLA can be configured on a per-table basis. The delay would specify by how much time the table can be behind real time. If a query comes in that would query data from that blackout period where data is only available in the raw table, but not yet guaranteed to be in the rollup table, the incoming query can be split into two using the SplitRollupQuery class. It wraps a query that queries the rollup table until the last guaranteed to be available timestamp based on the SLA; and one that gets the remaining data from the raw table. * Extract an AbstractQuery Extracts an AbstractQuery from the TsdbQuery implementation since we'd like to reuse some parts of it in other Query classes (in this case SplitRollupQuery) * Extract an AbstractSpanGroup * Avoid NullPointerException when setting start time Avoids a NullPointerException that happened when we were trying to set the start time on a query that would be eligible to split, but due to the SLA config only hit the raw table anyway. * Scale timestamps to milliseconds for split queries Scales all timestamps for split queries to milliseconds. It's important to maintain consistent units between all the partial queries that make up the bigger one. * Fix starting time error for split queries Fixes a bug that would happen when the start time of a query aligns perfectly with the time configured in the SLA for the delay of a rollup table. For a defined SLA, e.g. 1 day, if the start time of the query was exactly 1 day ago, the end time of the rollups part of the query would be updated and then be equal to its start time. That isn't allowed and causes a query exception.
* Allow end_time = start_time to be able to query and delete a single datapoint. * Updating tests to support start and end time being the same Co-authored-by: Øyvind Matheson Wergeland <oyvind@wergeland.org>
* Start of new HTTP API. * Start of new "is this table available across all/some regions" API. * Finish plausibly OK implementation of checking table availability. * Finish plausibly ok Telnet/HTTP status endpoint. * It compiles. * Minimal live test works (for success case). * Set a timeout on RPC queries. * Unit test for the HTTP status RPC query. * Start tests for checkNecessaryTablesAvailability(). * Checkpoint. Not working, probably because of old Mockito. * Fix API usage. * Got the test to pass. * More unit tests.
* added tsdb_list_running_queries.py * updated error handling * updated tsdb_list_running_queries.py * various cleanup, removal of old code etc * switched to custom ConnectionException * added comment for now_length Co-authored-by: Hari Sekhon <Hari.Sekhon@gresearch.co.uk>
…roup_by filters act as group_by filters
The / at the end of url makes url in this form https://repo1.maven.org/maven2/com/pythian/opentsdb/asyncbigtable/0.3.0//asyncbigtable-0.3.0-jar-with-dependencies.jar but that "//" in the url makes a 404!
johann8384 force-pushed the next branch 2 times, most recently from a47781e to e6dd3f3 Compare This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters. Learn more about bidirectional Unicode characters
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
18 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Fixes #758