88ce311b37
Remove all WAL-related files when loading new DB
1 year ago
b931e387be
First pass at removing in-memory support
1 year ago
f06e547ab5
Less disk IO during load
1 year ago
f5517c0f66
Correct comment
1 year ago
9bdf048975
Better naming of temp backup file
1 year ago
b5cb889e1e
Auto-restore doesn't require ready() check
1 year ago
b7633143cc
Restore from file uses chunking
1 year ago
5ca5ce342e
Support internode LoadChunk()
1 year ago
1c8e52c5f7
Test_SingleNodeLoadBinaryFromReader passes
...
That's good, but much more extensive testing needed, and this needs to
replace the existing load path -- and then we move to 8.0.
1 year ago
93a752ae0e
Better variable names
1 year ago
1c6f691ff8
Integrate chunking - WIP
...
I need more sophisticated dechunker management.
1 year ago
13d47b24d8
Store can load from an io.Reader
...
This may be helpful for chunked-loads.
1 year ago
a40c43725f
More logic reversal
1 year ago
f81aad8fd3
Revert logic for clarity
1 year ago
21dff4a68c
Remove on-disk-startup control
...
With the move to WAL and "synchronous mode" to OFF, on-disk startup
times are very close to in-memory. There is no need for this control
anymore and it complicates the start-up code.
1 year ago
f7ced46377
Count number of DB stats errors
1 year ago
36fa7f2afd
Detect DB stats error and log
1 year ago
7ce2c6b99a
Fix comment
1 year ago
7432bceca3
Since we're passed a ReadCloser, close it
1 year ago
4342650c33
Fix RecoverNode
1 year ago
ad73779285
Periodically record actually applied index
...
This ensures the system doesn't apply uncommitted log entries to the
FSM, if it runs the in-memory startup. There have been no reports of
this in the field -- it is an edge case -- but this removes any chance
of it.
1 year ago
abfe056b4c
Remove any -wal and -shm files on shutdown
...
If in WAL mode, open-and-close again to remove the -wal file. This is not strictly necessary, since any on-disk database files will be removed when rqlite next starts, but it leaves the directory containing the database file in a cleaner state.
1 year ago
b2f28d1181
Better WAL control
1 year ago
2f7672d897
Enable WAL mode
1 year ago
585f9c9c05
Use Snapshot decoder in Store
1 year ago
c87931c45c
Factor Snapshot creation into own module
1 year ago
3f462eecfa
Remove all on-disk SQLite files on startup
1 year ago
7e1b8d12d7
Add full WAL support to DB layer
1 year ago
eed215c664
Only 1 Snapshot needs to be retained
1 year ago
b9253e6b41
Improve logging in Store
1 year ago
4a7ac2850f
rqlite doesn't (yet) support WAL mode SQLite data
1 year ago
92bc513242
Move FSMSnapshot to own source file
1 year ago
7badba4a7a
Try fast paths
1 year ago
76ba50ca86
Log last (if any) WaitForLeader error
1 year ago
ef403cdcc6
Fix up WaitForLeader
...
Add dedicated error codes too.
1 year ago
c05e091cd1
Add WaitForRemoval
1 year ago
c23d3e60b1
Reuse existing consistency level
1 year ago
18b76ebd4f
go fmt
1 year ago
2f8d4e95e0
Update store.go
1 year ago
eea101a1d5
Fix duplicate import
1 year ago
f7cdbdea68
Handle Command_COMMAND_TYPE_EXECUTE_QUERY in Store
...
Works! Store.Request() needs unit testing however.
1 year ago
b2b0b188e9
Factor out functions
1 year ago
89f472743b
Start adding Store.Request()
1 year ago
7680483793
Tighten up remove
1 year ago
e64d78b809
Support removing self from cluster on shutdown
1 year ago
02b6f7df61
selfLeaderChange is not just about restores
1 year ago
159b127101
Clean up comments
1 year ago
0c6be22478
Close auto-restore file
1 year ago
1fa7cf73a7
Log any failure to delete auto-restore file
1 year ago
1941c947c7
Fix comment
1 year ago
466c757dcf
Add long-missing init call to Store stats
1 year ago
ee55a35c03
SetRestorePath unit testing
1 year ago
e7919a906f
Avoid the race during testing
1 year ago
31467482a3
Store-level multi-node auto-restore testing
...
This test ensures that auto-restoring works as expected when
Bootstrapping is enabled.
1 year ago
bf1e1cb1cd
Add first tested version of Store auto-restore
1 year ago
5c2e8e7ad6
Add channel which gets signal on becoming Leader
1 year ago
833d9d3a6f
Perform older checks first
1 year ago
5e3f294dfa
Only check for a ready store if making changes
1 year ago
2ce3c5f34c
Check Store is ready in more places
1 year ago
57ca03014b
More integration and testing of Store readiness
1 year ago
716f84ff28
Store supports registering "ready" channels
1 year ago
372e4dd4a9
Store implements simpler Provide() interface
1 year ago
5fcc222f73
Better Store comment
1 year ago
ae97060f2c
Implement the DataProvider interface on the Store
1 year ago
519df067cc
Check for more possible errors with peers
...
https://github.com/rqlite/rqlite/issues/1217
1 year ago
35df643a0b
Remove unneeded type assertions
1 year ago
1794724ef0
Remove unused code and fix error handling
1 year ago
93b3308502
Skip any files that disappear during walking
1 year ago
ff001bc5a8
Migrate to a Protos data model for Notify Requests
2 years ago
a782513a11
Migrate to a join-request proto
2 years ago
fdf5fce053
Update store.go
2 years ago
2730fa06e5
Empty config means not a voter
2 years ago
5636febf9a
Add IsVoter() and add to status/ output
2 years ago
defde21281
Merge remote-tracking branch 'origin' into ignore-join-explicit-if-state
2 years ago
90b4dc0745
Silently ignore unchanged node self-joins
2 years ago
54d61c7866
Update store.go
2 years ago
a2899a1cdd
Support stepping down before shutting down
2 years ago
0702e39ace
Add ResetStats
2 years ago
4da5b6322f
Log message could be confusing with remote remove
2 years ago
98575d727b
Start adding remote remove node
...
More testing, including unit testing, required.
2 years ago
bfc35eaf7c
Simplify reap command line flags ( #1118 )
...
Simplify reap command line flags
2 years ago
220424c7d1
Simple refactor, better Go style
2 years ago
07620cab41
Reap nodes ( #1114 )
...
Support automatic reaping of nodes
2 years ago
e158ae8a72
Start HTTP server ASAP
2 years ago
0e8e266f96
Merge pull request #1090 from rqlite/fix-remote-load-errors
...
Fix remote-load error handling
2 years ago
893fbb0249
More store-level load operation logging
2 years ago
65cd556549
Fix remote-load error handling
...
Error from remote node was being dropped.
2 years ago
5f33bffa0f
More logging and resolution checks
2 years ago
b0f17b86c7
Notified node checks address resolution
...
If a non-resolvable address is passed to the Store, and then Store then
adds that to the config, it can result in a cluster that seems to have
insufficient members (since the remote node can't be contacted). This
may trigger a new leader election.
2 years ago
47d79c859a
Log database backup time
2 years ago
9fc6ea16e9
Even better defer sequence
2 years ago
555883d1a2
Correct order of deferred calls
2 years ago
54aec1ca33
Tighten up backup code
2 years ago
31bba4c577
Move to Backup proto model
2 years ago
dc2b9f5b6f
Ignore freshness when serving queries on Leader
...
It doesn't make sense to check the last leader-contact time when the
node itself is the leader. Any read, when served by the leader, is going
to be within the freshness bound.
Fixes https://github.com/rqlite/rqlite/issues/1048
2 years ago
89b132ac26
Fix Store-level unit tests
2 years ago
b03ed99eda
Update comment
2 years ago
304124952c
Add Store level changes
...
Needs unit testing, and HTTP testing won't compile.
2 years ago
2729a17279
Move some functions to standalone form
2 years ago
9783ef24b1
Switch to error on self-join attempt
3 years ago
a524ed4979
Ignore join request if node ID is same
3 years ago
e4fed0cee3
Add bootstrap-expect ( #974 )
...
Add cluster-boostrap
3 years ago
ad7986630d
Better command-line flags
3 years ago
ce099087ee
Control whether bbolt sync freelist to disk
...
When true, skips syncing freelist to disk. This improves the database write performance under normal operation, but requires a full database re-sync during recovery.# Please enter the commit message for your changes.
3 years ago
490f654f36
Move to etcd's fork of BoltDB
...
bboltdb is the version the Consul now runs.
3 years ago
4aea326959
Consul and etcd autoclustering ( #957 )
...
This PR introduces new node-discovery integration with Consul and etcd. By using one of those systems with rqlite, automatic clustering of rqlite is much easier.
3 years ago
ca556044e3
Merge branch 'master' of github.com:rqlite/rqlite into refactor-main-bootstrap
3 years ago
2ffb29394f
Fix typo
3 years ago
f5c03d4b60
Try simple refactor of bootstrap
3 years ago
769557db7e
Better variable name
3 years ago
3bfcbffc37
Fix typos and Grammar issues
3 years ago
33bee34f89
More Store restart testing
3 years ago
d74c05c2db
Support disabling in-memory init of on-disk
3 years ago
003301aee6
Minor fixes to log messages ( #920 )
...
Minor fixes to log messages
3 years ago
0b621cbec8
More logging tweaks
3 years ago
5ce51bc2ba
Better comments
3 years ago
05534cc94c
Don't use Windows file paths as IDs
3 years ago
f9f35bf95c
Dump file
3 years ago
7093cb25d3
Fix race in test
3 years ago
9fb114c07f
Recover test after snapshot
3 years ago
519a91e635
Actual recovery node test passes on Linux
3 years ago
1b158d09fb
Simple unit test of node-recovery
3 years ago
1abd3007ca
Rename peers.json after successful recovery
3 years ago
cc1de72796
Ensure raft/ exists inside data directory
3 years ago
abccc6ced3
Better function name
3 years ago
d9ad3c080a
Add untested recover-node functionality
3 years ago
32239b5053
Move new function to earlier per convention
3 years ago
4ab72a8194
Refactor to make recover-cluster code easier
3 years ago
5dd4aeed1e
More Snapshot metrics
3 years ago
38eef8ee82
Merge pull request #890 from rqlite/more-abl
...
ABL
3 years ago
dd71c07c73
ABL
3 years ago
0e34dc49ca
Add BoltDB stats
3 years ago
d6be09ba65
Update Snapshot comment
3 years ago
5d48a18950
Check if slice is empty in addition to being nil
3 years ago
bb002bb81b
Block query transactions during serialize
3 years ago
a6da8d474f
Ensure (any) directory for on-disk-path exists
3 years ago
1f33dbb2cb
Fix up Store-open logging
3 years ago
105a894fe5
Support explicit SQLite path at Store level
3 years ago
a6346ba9d4
Fix unit test -- start muxes
3 years ago
dd692ecf4f
Add gauge-like metric for Snapshot times
3 years ago
9d87a15194
Small simplication
3 years ago
43e60a0469
Perform leader check first for STRONG query
...
Don't wait until Raft layer returns it. There is a small chance of a
false leadership detection, but the cluster would have to be in flux
anyway for that to be the case. Execute does this already.
3 years ago
fd98114839
HTTP-level testing passes
3 years ago
3580e341a8
Store-level testing passes
3 years ago
deb4dabf2e
Thread through foreign key constraint controls
3 years ago
0536bc3c1c
ABL
3 years ago
74d0780a5f
Merge pull request #845 from rqlite/sqlite-options
...
Add SQLite compile-time options to status output
3 years ago
39d0f2ad4a
Merge branch 'master' of github.com:rqlite/rqlite into sqlite-options
3 years ago
fc76327b46
Use FSM index in tests
3 years ago
b8d72f3fc4
New DB and FSM indexes to track state
3 years ago
ea723ee830
Ping databases on open to ensure full availability
3 years ago
696bac655f
Merge branch 'master' of github.com:rqlite/rqlite
3 years ago
f03900b282
Log type of database on start-up
3 years ago
b83424484d
Enhance SQLite database status
3 years ago
6edff5ba24
Deal with SQLite connection "caching"?
...
https://github.com/mattn/go-sqlite3/issues/966
3 years ago
6dc9583dab
Implement concurrent reads and writes
...
Disk doesn't support it yet, as it has not moved to WAL mode. Perhaps
wait for next release to do that.
3 years ago
e0da96dd55
Remove support for specifying SQLite DSN
...
This is a breaking change. However this feature hasn't been tested and
allows end-users to break the system too easily. Low-level control over
the SQLite database is better done with PRAGMA commands where possible.
3 years ago
8693694d14
Add current SQLite journal mode to status output
3 years ago
503825dbaf
Remove Abort functionality
...
No longer relevant since each Exec and Query gets its own connection.
3 years ago
561bb47a14
Port DB layer to sql/database
...
Empty statements now cause panic.
3 years ago
5f5f7dbbea
Sync use of DB connection
...
Possible, albeit simple, fix for https://github.com/rqlite/rqlite/issues/830
3 years ago
67b0368288
It's create in memory, not open
3 years ago
0a55f0bed1
It's create, not open
3 years ago
01e2403015
Count various Join scenarios
3 years ago
4591e2df35
Don't ignore -join even if previous state exists
3 years ago
f0417f2241
Correct typo in comment
3 years ago
58b567c27a
Only check non-voting nodes by default
3 years ago
4fd5555c41
ABL
3 years ago
f939d33fe5
Remove Metadata concept from system
...
System testing passes.
3 years ago
b2bcdb584d
Tweak Store API
3 years ago
b579e5c59f
Shut down Raft first
4 years ago
1dc156b965
Add total Raft directory size to Store stats
4 years ago
fd4b76a419
Convert Raft stats to numbers where possible
4 years ago
125074ab6e
Only shutdown BoltDB when Raft is done with it
4 years ago
3ff13cda74
More system testing
4 years ago
51c745ea23
More in-mem and on-disk system test
4 years ago
ae16a65d8d
Handle snapshots with database data
...
This shouldn't happen normally, but technically could.
4 years ago
f8e4795d44
Merge branch 'master' into test-snap-restore-more
4 years ago
c2f40500a2
Close Bolt on Store close
4 years ago
e4c60bed7a
Merge branch 'master' into test-snap-restore-more
4 years ago
ca1f8e1752
Whitebox testing of Noops on Store
4 years ago
0b84667b45
Fix on-disk create in Apply
4 years ago
a8559b49e8
Merge pull request #754 from rqlite/noop-store-test
...
Support Noop commands in Raft Log
4 years ago
824dab5536
Doesn't work because Store can't be re-opened
...
Due to network issues.
4 years ago
d31105e0f0
Add missing return
4 years ago
ed080fe9d0
Close database before Restore
4 years ago
9c53b0f8b7
Consolidate database creation functions
4 years ago
c782b43f2d
Build on-disk databases in-memory first
...
With this change, rqlite nodes running in "on-disk" mode build the database in memory first, and move it to the disk just before the Raft system starts. This means that on-disk nodes now initialize almost as quickly as in-memory nodes.
4 years ago
33aa4522cf
Better timing from first to last log
4 years ago
a4185172da
Time snapshot, restore, and startup times
4 years ago
80e0db9f52
Make logic easier to read
4 years ago
9ea54b608f
Use DeserializeInMemoryWithDSN for in-mem restore
4 years ago
ef1f0f1d57
Sync queries and restores
4 years ago
0f1352fb01
Remove duplicate pre-restoration Close()
4 years ago
ce7a3a6912
Merge branch 'master' into no-disk-during-restore
4 years ago
0b3a41bd3f
Clearer comment
4 years ago
535d690d7c
Tigthen up Store-level locking
...
Previous locking didn't look quite correct, and with the advent of the
Serialize functionality, it needs to be more precise.
4 years ago
0d0bb3f44a
Close database before restoration
...
This change also adds better logging, so the restoration process can be
better observed.
4 years ago
eb01e8dc9d
Skip temp file when restoring in-memory database
...
This saves a read and write to disk, the size of the SQLite database.
4 years ago
db64f4c923
Don't use temp file during database snapshot
...
Snapshotting the database does not now need to write to disk, and copy
the SQLite database straight to RAM, for handoff to the Raft subsystem.
4 years ago
93a78f09a0
Add call to Serialize function
4 years ago
d560356d8d
Fix spelling mistakes
4 years ago
8576ed3cfd
Remove unused byte slice address
4 years ago
bb31aef20a
ABL
4 years ago
9b0852e711
Implement policy over trailing logs
...
Don't allow control over trailing logs directly, just implement a policy
that scales the number by 1.25 relative to Snapshot threshold. Without
doing this it may be that setting the snapshot threshold lower doesn't
actually affect SQLite start up times, due to log application. Perhaps
this really needs exposure at the command line, but the command line is
starting to get complicated.
4 years ago
8cb88d7744
Count number of legacy commands unmarshaled
4 years ago
4f0c3092b3
Better command decoding panic message
4 years ago
7780a0987b
Simplify variable declaration
4 years ago
e52b4c58c3
Factor out binary value functions
4 years ago
011bc00784
Add SQLite database page-centric size to status output
4 years ago
6d77d904e7
Compress SQLite database in Raft snapshot
...
Before Raft truncates the log, it needs to retrieve a snapshot of the SQLite database to represent the log entries that are removed during that truncation. Raft then writes that snapshot to disk. To reduce both disk IO and disk space this change compresses that SQLite snapshot before it is written to disk, and decompresses it during a restore operation.
This change also deals with older SQLite snapshots, which were not compressed.
4 years ago
d5322a1909
Ignore join request if node already part of cluster ( #711 )
...
rqlite used to work like this, but suffered a regression due to a change in how Hashicorp Raft worked. The manner it changed in was not public, so relying on it was always fragile.
4 years ago
6575de779d
Use Protobuf for encoding Raft Log commands
...
This PR changes Raft Log Entry encoding from JSON to Protobuf. Furthermore, larger Raft commands (which can result from batching SQL statements, or individually long SQL statements) are compressed before encoding.
This primary reason for this change is to reduce IO load since that is one of the largest performance bottlenecks. It will also reduce internode traffic.
Legacy JSON-encoded commands are still handled by this code, so this change is backwards-compatible with previous releases in the v5 series.
4 years ago
b4867713e4
Better variable name, Query -> SQL
4 years ago
b6ebdf3ab1
Support setting Leader Lease timeout
4 years ago
8a4a5c4901
Improve comments re opening of Store
4 years ago
f961bb4e44
Fix typos and comments
4 years ago
1480d8898f
Handle Raft commands missing Parameterized values
...
This allows this code to run with Raft logs from previous versions.
4 years ago
ac862e4091
Add system test for parameterized statements
...
Fix some bugs along the way too.
4 years ago
212d4e5c86
Store and Raft layer support parameterized queires
4 years ago
d840adc2a9
Convert Store to use new DB Statements
4 years ago
f81ef9024e
Add Raft log size on disk to status
4 years ago
c4a89ede8d
Add helper function for Raft log size
5 years ago
487c0e3f8c
Set Snapshot Interval to non-zero value
...
Snapshotting was *never* functional due to the zero interval value.
5 years ago
9c4151c37b
move comment down
...
Change-Id: I2f8c435adab5e1c46da2e6b91205ad4df850decc
5 years ago
9c295ca704
check level before freshness
...
Change-Id: I6d736ad09eac9f003f087adaa9c1f51691fd5c69
5 years ago
536cac4fdf
Allow log level control for Raft module
...
Default log level is INFO.
5 years ago
abb7772105
Support leader freshness with None consistency
...
This change allows the read request to specify the maximum time the node
receiving the request may have last heard from the cluster leader. It
only applies to a read consistency selection of None.
5 years ago
81f5f6faf1
Add support for non-voting nodes
...
A non-voting node doesn't participate in Raft consensus, but does
subscribe to the committed log entries originating with the leader.
This means a non-voting node keeps up-to-date with the state machine,
without impacting write-latency. These non-voting nodes can provide
read scalability for the cluster.
5 years ago
9d595e2ead
ABL
5 years ago
e0cf01cefc
Protect against possible nil pointer
5 years ago
a23af88235
Write backup and dump files to a writer
...
This avoids making an extra copy in RAM. Port of original PR at
https://github.com/rqlite/rqlite/pull/461/ .
5 years ago
f57ace7da2
Broadcast Store meta via standard consensus
...
With this change the cluster metadata (arbitrary key-value data associated with each node) is now broadcast across the cluster using the standard consensus mechanism. Specifically the use case for this metadata is to allow all nodes know the HTTP API address of all other nodes, for the purpose of redirecting requests to the leader.
This change removed the need for multiplexing two logical connections
over the single Raft TCP connection, which greatly simplifies the
networking code generally.
Original PR https://github.com/rqlite/rqlite/pull/434
5 years ago
e376812907
Add method to Store that returns leader ID
...
With the advent of Raft IDs, this distinction matters.
5 years ago
8a0e1ce8f0
Move to Hashicorp Raft v1
5 years ago
7352c376c7
Don't ignore Join error, actually return it
...
Clearly a typo.
5 years ago
085d2d23d8
ABL
5 years ago
0e69be3949
Store layer supports generating SQL format backups
...
Port PR453.
5 years ago
0471877f26
Abort transaction if load operation fails
5 years ago
ccca1ecfa9
Fix DSN key in Store status output
5 years ago
d47808a26b
Add expvar stats to Store
5 years ago
2e91858e1e
Allow Raft election timeout to be set
6 years ago
6937a377ae
Allow control of shutdown-on-remove
...
Defaults to leaving Raft run if the node is removed.
6 years ago
efec4d4e42
Return redirect if remove on follower
...
Fixes issue #391 .
7 years ago
b5b2cb21db
Log Store directory path on startup
7 years ago
6bbdc88c1d
ABL
...
Always be linting.
7 years ago
2a1b6c65c3
New QueryRequest and ExecuteRequest types
...
The Query() and Execute() functions on the Store now take a complex type
that encapsulates the statements, and associated parameters. This will
make it easier to add more control parameters when making requests.
7 years ago
303be73dcb
Add missing Store params to diagnostics
7 years ago
2617962f5f
Typo
8 years ago
4cb5035172
remove store.readPeersJSON
8 years ago
Philip O'Toole
chermehdi
wangfenjin
WanliTian