d627634991
Ensure internal retention policy exists
2015-03-15 14:32:46 -07:00
ae3b3d5252
Rename "metrics" to "stats"
2015-03-15 14:32:46 -07:00
5240151517
Actually start self-monitoring if requested
2015-03-15 14:32:11 -07:00
bf28720c6e
Add "name" to stats object
...
This "name" becomes the Measurement name. In addition the implementation
for "SHOW STATS" has been re-instated.
2015-03-15 14:32:11 -07:00
6867d806de
Use Inc() shorthand for Stats
2015-03-15 14:32:11 -07:00
f730257ee4
Implement self-monitoring
2015-03-15 14:31:16 -07:00
228e238436
Remove unnecessary err variable
2015-03-15 14:31:16 -07:00
8825580500
Track server stats
2015-03-15 14:31:16 -07:00
b842bf2446
Merge pull request #1952 from dullgiulio/delete-no-panic
...
return error on delete statement, closes #1933
2015-03-15 10:16:42 -07:00
06d839223e
Integration test delay.
2015-03-14 14:31:25 -06:00
41d357ac04
Fixes based on code review comments.
2015-03-14 14:07:09 -06:00
53dbec8232
Add config notifications and increased test coverage.
2015-03-14 13:36:06 -06:00
2cc2077c3d
do not accept empty database name in drop
2015-03-14 21:31:44 +02:00
38bb2cf62e
do not allow empty database name, closes #1950
2015-03-14 15:39:22 +02:00
a4f96a3b9f
return error on delete statement, closes #1933
2015-03-14 11:14:32 +02:00
4b9a93d924
Merge branch 'master' of https://github.com/influxdb/influxdb into stateless-broker
2015-03-12 15:46:04 -06:00
d82d040e59
Create databases for Graphite at higher level
...
By doing this the SeriesWriter interface stays focused and just has
methods for writing data. Its name then remains coherent.
2015-03-12 13:33:25 -07:00
7880bc2452
Add zero length data checks.
2015-03-12 12:12:26 -06:00
7ab19b9f91
Merge branch 'master' of https://github.com/influxdb/influxdb into stateless-broker
2015-03-12 11:59:55 -06:00
ddc6e119ab
Add default column to SHOW RETENTION POLICIES
2015-03-12 11:07:51 -06:00
8655ed3d91
fixing default retention policy creation
2015-03-12 10:01:26 -06:00
5f6bcf523f
Fix broker integration bugs.
2015-03-11 12:00:45 -06:00
982e59a416
Merge pull request #1906 from influxdb/show-servers
...
Add show servers to query language
2015-03-11 10:38:11 -06:00
f6121a9b63
fixing edge case for infinite retention policy
2015-03-11 10:14:53 -06:00
66115f95ca
Merge branch 'master' of https://github.com/influxdb/influxdb into stateless-broker
...
Conflicts:
cmd/influxd/run.go
messaging/broker.go
messaging/client.go
messaging/client_test.go
messaging/intg_test.go
server.go
tx_test.go
2015-03-10 16:27:37 -06:00
5e65a07b31
retrieve list of data nodes
2015-03-10 15:47:34 -06:00
27e9132796
Integrate stateless broker into remaining packages.
2015-03-10 14:53:45 -06:00
db067b6a2d
move constant
2015-03-10 12:24:26 -06:00
91487cc8dd
update comment
2015-03-10 10:52:23 -06:00
b8fb723967
Enforce retention policy minimum (currently 1 hour)
2015-03-10 10:49:54 -06:00
c150a8c8bc
run all create shards groups through same methods
2015-03-09 18:54:37 -06:00
7495e6eb1c
fix comments, some renaming, adding rlock
2015-03-09 18:05:39 -06:00
b6e7210f47
close channel
2015-03-09 17:36:50 -06:00
67f03ef26b
Pre-create shard groups
...
1) Pre create shard groups
2) Calculate and save the ShardGroupDuration for Retention Policies. This should never change even if they alter a retention policy in the future
2015-03-09 17:30:17 -06:00
abf664c18d
Merge pull request #1895 from influxdb/create_default_rp
...
Create retention policy by default
2015-03-09 16:26:41 -07:00
f72313edd9
Support retention policy auto-creation
2015-03-09 16:24:03 -07:00
5f5c6ca297
Integrate stateless messaging into influxdb package.
2015-03-09 15:47:41 -06:00
713ca4b831
Merge branch 'master' into stateless-broker
...
Conflicts:
messaging/broker.go
2015-03-09 12:27:14 -06:00
1db4851e49
RLock Server when checking retention policies
2015-03-09 10:56:59 -07:00
9b63eed767
Merge branch 'master' of github.com:influxdb/influxdb into 1753
...
Conflicts:
CHANGELOG.md
2015-03-06 12:37:15 -08:00
771ef1cec3
Don't Panic on Missing Dirs
...
Issue: 1753
Figure out what's missing, create it, and keep going.
2015-03-06 12:11:02 -08:00
76c255a3c8
Add trace-level logging to server write path
2015-03-05 17:35:57 -08:00
93b56ca058
Make sure queries for missing measurements return a `200 OK` even when using a `GROUP BY *`.
2015-03-05 17:27:46 -08:00
77d8986626
Don't delete data with infinite retention
2015-03-05 15:22:36 -08:00
1bbf154f2a
Removing replicas and subscriptions from broker.
2015-03-01 22:32:20 -07:00
ef262171d6
trim command line before parse it
...
This can avoid some errors caused by an additional space at the end of
the command line.
For example: 'use db ' will be trimed to 'use db' and parsed correctly.
Signed-off-by: Kai Zhang <zhangk1985@gmail.com>
2015-03-01 10:44:20 +08:00
a7c0d71d9c
Merge pull request #1770 from kylezh/dev
...
Fix error when alter retention policy
2015-02-27 20:55:14 -08:00
a26f11f489
fixes based on @dgnorton's feedback
2015-02-27 01:09:38 -05:00
52f257e349
Make show series return IDs and work with limit and offset.
...
Fixes #1765
2015-02-27 01:09:38 -05:00
9f8639ded8
Fix error when alter retention policy
...
+ fix error when execute ALTER RETENTION POLICY bar ON foo DURATION 1h
+ add test to verify the fix
Signed-off-by: Kai Zhang <zhangk1985@gmail.com>
2015-02-27 09:54:39 +08:00
0ab43cc8a5
adding better error message to return
2015-02-26 17:10:09 -07:00
499a5dcb33
show series should only return for measurements with series data
2015-02-26 17:10:09 -07:00
66c5750726
return early if no measurement. clarify comments
2015-02-25 12:56:23 -07:00
fb2f7416ed
drop shard data
2015-02-25 12:42:23 -07:00
a151fe64d4
deleting data from metastore. wip
2015-02-25 12:42:23 -07:00
0ad76c637d
broadcastig drop measurement
2015-02-25 12:42:23 -07:00
079f64338b
wiring up droping series, wip
2015-02-25 12:42:23 -07:00
eb0480ebdf
Perform retention checks again UTC time
2015-02-25 10:55:48 -08:00
e7e04cdeb0
fix #1717 : code review chng & update CHANGELOG.md
2015-02-24 14:35:22 -05:00
95a5f2588e
fix #1717 : all points must have at least one field
2015-02-24 14:09:10 -05:00
86972788d3
Merge pull request #1632 from influxdb/wildcard_group_by
...
Implement comprehensive query wildcard re-write
2015-02-24 00:13:37 -05:00
41ad49f1c1
Add wildcard detection method to SelectStatement
2015-02-23 19:09:55 -08:00
84de4ad56c
Add wildcard rewrite functionality for SELECT
...
This code, when given a SELECT statement, returns another SELECT
statement such that all "query" and GROUP BY wildcards have been
replaced with all Measurement fields and tag keys respectively.
2015-02-23 19:09:55 -08:00
86b91ead94
"SHOW TAG VALUES" now outputs series per tag key
...
This change tightens up the type used for "Series" so the pre-existing
sort method can be used.
2015-02-23 17:15:55 -08:00
2f0fe901e3
Rename "Values" to "Fields" for writes
2015-02-23 15:06:56 -08:00
2cc23d3c66
Merge pull request #1529 from influxdb/stats-reporting
...
Bring back stats reporting.
2015-02-23 12:57:49 -08:00
f647cdc031
Add read locking and inline data collection.
2015-02-23 12:37:17 -08:00
8e086a447b
Change delete to drop for [dD]ropDatabase
2015-02-23 13:14:18 -07:00
20a5d0b932
Rename ids for clarity.
2015-02-23 10:42:39 -08:00
c059e21974
Also track number of databases.
2015-02-23 10:42:39 -08:00
20e7f2a750
Some small cleanups.
2015-02-23 10:42:39 -08:00
e45611cde3
Pull in cluster ID for reporting.
2015-02-23 10:42:39 -08:00
ddf1332ff8
Add reporting of series and measurement counts
2015-02-23 10:42:39 -08:00
ff1c61c5c1
Only report if ReportingDisabled is set to false.
2015-02-23 10:42:39 -08:00
2ac2538ed3
Don't block for stats reporting.
2015-02-23 10:42:39 -08:00
c366fcbbf6
Update stats reporting to just POST for now.
2015-02-23 10:42:38 -08:00
3e727ac01e
Change "Rows" to "Series" in API output
2015-02-22 21:21:49 -08:00
9ab1221a35
Move Raft command types to own file
...
server.go is getting long, so start breaking it up a little.
2015-02-22 11:50:43 -08:00
8637a10b6a
start with seriesByMeasurement
2015-02-21 22:18:25 -07:00
29910c3c99
better transaction batching for drop series
2015-02-21 21:48:18 -07:00
d5ac049812
remove bogus method
2015-02-21 20:20:36 -07:00
d35e2a92b1
locking is now done in process messages
2015-02-21 19:14:00 -07:00
c575ee9a04
fixing a fat fingering rebase
2015-02-21 19:13:01 -07:00
2752ada587
moving more responsibility to database.dropSeries
2015-02-21 17:21:27 -07:00
6114c8138e
refactoring drop series
2015-02-21 17:21:27 -07:00
61352f78b6
pass all series IDs to Server.DropSeries at once
2015-02-21 17:18:50 -07:00
9a6e29ece8
collect all series IDs then unlock
2015-02-21 17:18:50 -07:00
5ab3731159
delete -> drop for series naming
2015-02-21 17:18:50 -07:00
697e9721dd
fixing rebase conflicts
2015-02-21 17:10:12 -07:00
0d6b7616de
accidentally lost this in my rebase
2015-02-21 17:10:11 -07:00
495cedb53b
move where we unlock
2015-02-21 17:10:11 -07:00
4948b62784
first working version of drop series. more testing to come
2015-02-21 17:10:11 -07:00
8306057179
actually drop shard data
2015-02-21 17:10:11 -07:00
6d5be3da25
wip. compiles now. tests next
2015-02-21 17:10:11 -07:00
7cfc2a5601
wip
2015-02-21 17:10:11 -07:00
6370a4ef2b
wip in progress for actually dropping the series data
2015-02-21 17:05:08 -07:00
4e5b6e93ff
Merge pull request #1672 from influxdb/data-node-recovery
...
Add index tracking to metastore.
2015-02-21 15:34:54 -07:00
4dbd1541f7
Add index tracking to metastore.
...
This commit adds tracking of the index to the metastore so that data
nodes will ignore replayed indexes when they are restarted. This
does not include index persistence for the shards.
2015-02-21 13:46:16 -07:00
9bff9cdeee
issue #1431 : wire up regex queries
2015-02-21 15:38:35 -05:00
9c4174a006
Simplify locking in WriteSeries()
...
In addition, memomize the Field codecs.
2015-02-20 14:26:12 -08:00
612ef1fa08
Write batch in a single BoltDB transaction
2015-02-20 11:28:07 -08:00
bd4352c77d
Remove unused shardsBySeriesID
...
Its presence is making Bolt-level batching quite awkward, and since it's
not used, just remove it.
2015-02-20 11:28:07 -08:00
169409ac03
Move to slices for creating measurements
...
Using maps was resulting in unpredicatable ordering of columns and tags.
2015-02-20 11:28:07 -08:00
f5b2962d42
Incorporate inital code review feedback
2015-02-20 11:28:07 -08:00
5c61b7d8b8
Add new fields even when Measurement exists
2015-02-20 11:28:07 -08:00
0e3e223e84
Always ensure measurement exists in command
...
Unit tests need updating since some tests are no longer valid.
2015-02-20 11:28:07 -08:00
2585a9ea4f
Encode real batches for shards
2015-02-20 11:28:07 -08:00
ff765793e6
Implement applyCreateMeasurementsIfNotExists
2015-02-20 11:28:07 -08:00
25c3b104b7
WriteSeries now uses batching
2015-02-20 11:28:06 -08:00
0d3ab9fa1d
Create-measurement command in local function
2015-02-20 11:28:06 -08:00
3435da9626
Correctly initialize createMeasurement maps
...
Full unit tests added for happy paths.
2015-02-20 11:28:06 -08:00
55fbd7e42c
Batch Measurement creation
2015-02-20 11:28:06 -08:00
6fd7dd028a
Fix conflict with createFields and createContinuousQuery message types
2015-02-17 21:06:48 -05:00
36d16a3172
Update broker to have URLs for replicas
...
* Update the infludb broker to not need a server so that it'll work on non-data node servers
* Update messaging broker to keep track of connect urls for replicas
2015-02-17 21:02:53 -05:00
8fee14fbf8
Fixes based on PR feedback.
...
* Pull out magic numbers for running CQs in the broker
* Typo fixes in comments
* Update the process_continuous_queries endpoint to not have gzip
* Remove previous implementation of MeasurementNames in favor of new simpler one
2015-02-17 21:02:53 -05:00
edcc20bcab
Wire up show continuous queries
2015-02-17 21:02:52 -05:00
8d9bcdbc97
Finish wiring up basic version of CQs
2015-02-17 21:02:52 -05:00
fec6764b09
WIP: wire up continuous queries
...
* Fix run to work with CQ broker
* Fix CQ handler
* Fix SetTimeRange to use RFC3339Nano
* Fix the time range tests
* Fix the parser to parse for RFC3339Nano literals in addition to the other format
* Add logic for running CQs
* Remove duplicate WriteBufferSize default setting from config
2015-02-17 21:02:18 -05:00
b0d4b6da55
WIP: fix cq query times
2015-02-17 21:01:21 -05:00
f2d132b361
Implement continuous queries with group by intervals
...
* Update defaults on Broker for when to trigger CQ runs
* Add config settings for how often to calculate and recalculate CQ results
* Update Server to have CQ settings
* Update AST to fold removed time literals when setting time range
* Add periodic CQ running functionality to server
2015-02-17 21:00:24 -05:00
4217778a08
Add periodic triggering of continuous queries to broker.
...
* Create influxdb specific broker that will periodically start continuous query calculation. Will send to the same data node as long as it's up and responding
* Add endpoint in handler to start processing of continuous queries on a data node
2015-02-17 21:00:24 -05:00
5f14d5faf5
WIP: start implementation of continuous queries
...
* Update SelectStatement to have GroupByInterval
* Update parser to verify that CQs have GroupByInterval if Aggregated
* Implement CreateContinousQuery on Server
* Add ContinuousQueries to database
2015-02-17 21:00:24 -05:00
bacd500a7f
fix typo stuct to struct
2015-02-17 16:53:31 -05:00
e74fadfd53
Merge branch 'master' of github.com:influxdb/influxdb into support_all_value_types
2015-02-17 12:55:12 -08:00
ae81dff9f6
Use Field codec in debug code
2015-02-16 16:43:35 -08:00
6db74aa9bb
Implement codec's encode
2015-02-16 16:01:52 -08:00
8cdee6a997
Clarify use of ReadSeries()
2015-02-16 10:28:06 -08:00
dcdb4b77a1
Merge branch 'master' into support_all_value_types
...
Conflicts:
database.go
server.go
2015-02-16 09:21:10 -08:00
9db17a7e09
Only update Metastore if new fields were created
...
It is quite possible that multiple applyCreateFieldsIfNotExist messages
will sit unapplied in the Raft commit log, each carrying exactly the same
information about new fields. So during application of this message only
change the Metastore if the fields actually changed.
2015-02-15 16:58:29 -08:00
276c5ded5a
int maps to influxql.Number
2015-02-13 16:28:12 -08:00
9061a70953
Don't perform marshalling during field creation
2015-02-13 15:59:46 -08:00
31f65361e4
Fix up locking
2015-02-13 15:41:27 -08:00
75ebfaf3dd
Use "create fields" broadcast message
2015-02-13 15:41:27 -08:00
32faba7248
Add "create fields" broadcast message
2015-02-13 15:41:27 -08:00
c0655fb45d
Release RLock if database does not exist
2015-02-13 15:19:17 -08:00
0ca98ea6c2
Remove unmarshalValues
2015-02-12 16:07:25 -08:00
5b8c0748af
Remove marshalValues
...
Code now on Measurement type
2015-02-12 16:07:20 -08:00
bf9bfd6106
Error if field's type has changed
2015-02-12 16:06:37 -08:00
8f9f8051a4
fix #1579 : SHOW MEASUREMENTS format
2015-02-11 23:03:44 -05:00
19cd5509c0
Retention policy and database were reversed in log
2015-02-11 10:11:09 -08:00
8cb06c5b8e
Refactor retention policy enforcement for testing
2015-02-11 09:54:07 -08:00
ddb57b4569
Explain why shard deletion may fail
2015-02-11 09:54:07 -08:00
91f863258c
Close shard before deletion
2015-02-11 09:54:07 -08:00
65edd6b277
Detect shard groups due for deletion
2015-02-11 09:54:07 -08:00
6a20100d2f
Implement "delete shard group" command
2015-02-11 09:54:07 -08:00
37647cc6a6
Add function to return shard group by ID
...
If shard groups were stored in a map by ID, this lookup could be sped
up.
2015-02-11 09:54:07 -08:00
50ef1672cb
Add stubbed enable of retention policy enforcement
2015-02-11 09:54:07 -08:00
e4178520da
Close shards on server close.
2015-02-10 22:18:36 -07:00
d00718c3a8
Update retention policy before saving database
...
Without this change, new shard group information is not persisted in the
metastore and only when the next database-save took place would this
update be persisted. If the server shutdown before the update, the new
shard group would be lost.
2015-02-10 19:21:02 -08:00
c003b39611
Only need read-lock for path
2015-02-10 16:53:43 -08:00
Philip O'Toole
Philip O'Toole
Todd Persen
Ben Johnson
Giulio Iotti
Cory LaNou
Joseph Rothrock
Todd Persen
Kai Zhang
Paul Dix
David Norton
Frontend