Commit Graph

601 Commits

Author SHA1 Message Date
Matt Stancliff 76efe1225f Scripting: Fix regression from #1118
The new check-for-number behavior of Lua arguments broke
users who use large strings of just integers.

The Lua number check would convert the string to a number, but
that breaks user data because
Lua numbers have limited precision compared to an arbitrarily
precise number wrapped in a string.

Regression fixed and new test added.

Fixes #1118 again.
2014-06-10 14:26:13 -04:00
antirez 605303d9c6 Cluster test: set config epochs when creating clusters. 2014-06-10 16:56:43 +02:00
antirez 64e6d7af47 Cluster test: unit 03, check epochs correctness. 2014-06-10 15:46:36 +02:00
antirez 4fd35a28d3 Cluster test: better failover detection in unit 03. 2014-06-10 15:46:00 +02:00
antirez 92dcae8fe4 Cluster test: added show-redis-logs command in the interactive console. 2014-06-10 15:22:40 +02:00
antirez 1617c36c35 Cluster test: new unit 03, failover loop stress testing. 2014-06-10 15:05:52 +02:00
antirez 2ed450c322 Cluster test: 02 unit more reliable waiting for slave sync. 2014-06-10 15:01:52 +02:00
antirez 9865ce0e64 Cluster test: unit 02 should wait for failover. 2014-06-10 14:18:54 +02:00
antirez 6785b39eba Cluster test: check master -> slave role switch. 2014-06-10 13:54:45 +02:00
Salvatore Sanfilippo c7f93143f6 Merge pull request #1669 from mattsta/blpop-internally-added-keys
Fix blocking operations from missing new lists
2014-06-09 11:37:28 +02:00
Salvatore Sanfilippo e8ebd7d0e0 Merge pull request #1789 from yoav-steinberg/fix_eval_in_tests
Fix eval usage in tests to conform with eval semantics
2014-06-06 10:37:57 +02:00
antirez 134fd9eaf4 Tests for min-slaves-* feature. 2014-06-05 10:46:12 +02:00
antirez 3307db49bd Regression test for issue #1118. 2014-06-04 18:51:20 +02:00
antirez efbe173962 Cluster test: add tmp dir to Git repo. 2014-05-26 18:08:12 +02:00
Salvatore Sanfilippo 3e0aef1e2e Merge pull request #1775 from mattsta/fix-test-against-new-PID-format
Fix test framework to detect proper server PID
2014-05-26 17:56:58 +02:00
antirez 3495caf4fb Cluster test: catch FLUSHALL errors on node reset.
FLUSHALL will fail on read-only slaves, but there the command is not
needed in order to reset the instance with CLUSTER RESET so errors can
be ignored.
2014-05-26 11:00:11 +02:00
Matt Stancliff 6c16ecaaaa Fix test framework to detect proper server PID
Previously the PID format was:
[PID] Timestamp

But it recently changed to:
PID:X Timestamp

The tcl testing framework was grabbing the PID from \[\d+\], but
that's not valid anymore.

Now we grab the pid from "PID: <PID>" in the part of Redis startup
output to the right of the ASCII logo.
2014-05-23 13:54:29 -04:00
antirez d0566daeaf Cluster test: basic failover unit added. 2014-05-23 11:47:47 +02:00
antirez aa5dfb3c2c Cluster test: move basic read/write test into a procedure. 2014-05-23 11:41:50 +02:00
antirez a700bc74a8 Cluster test: more reliable 01-faildet unit.
Do things in a sequence that prevents failover during failure detection.
2014-05-23 11:40:34 +02:00
antirez ce37488919 Test: AOF test false positive when running in slow hosts.
The bug was triggered by running the test with Valgrind (which is a lot
slower and more sensible to timing issues) after the recent changes
that made Redis more promptly able to reply with the -LOADING error.
2014-05-22 16:05:03 +02:00
antirez 17b05afda9 Test: dump.tcl fixed for RESTORE new error msg. 2014-05-22 15:56:17 +02:00
Matt Stancliff 33f943b4cd Fix blocking operations from missing new lists
Behrad Zari discovered [1] and Josiah reported [2]: if you block
and wait for a list to exist, but the list creates from
a non-push command, the blocked client never gets notified.

This commit adds notification of blocked clients into
the DB layer and away from individual commands.

Lists can be created by [LR]PUSH, SORT..STORE, RENAME, MOVE,
and RESTORE.  Previously, blocked client notifications were
only triggered by [LR]PUSH.  Your client would never get
notified if a list were created by SORT..STORE or RENAME or
a RESTORE, etc.

Blocked client notification now happens in one unified place:
  - dbAdd() triggers notification when adding a list to the DB

Two new tests are added that fail prior to this commit.

All test pass.

Fixes #1668

[1]: https://groups.google.com/forum/#!topic/redis-db/k4oWfMkN1NU
[2]: #1668
2014-05-21 09:52:52 -04:00
antirez 7f772355f4 Regression test for issue #1764. 2014-05-20 16:20:16 +02:00
antirez a9e62ab9fa HyperLogLog regression test for issue #1762. 2014-05-19 15:44:04 +02:00
antirez 194b78050e Cluster test: better failure detection test and framework improvements. 2014-05-19 15:26:19 +02:00
antirez a7da78e472 Cluster test: failure detection initial tests. 2014-05-19 11:39:15 +02:00
antirez 4c04744734 Cluster test: proper initialization at unit startup. 2014-05-19 11:24:15 +02:00
antirez 0b7aa26164 Cluster test: added function assert_cluster_state. 2014-05-14 15:21:57 +02:00
antirez 27ca133d35 cluster.tcl: fix redis links leak in refresh_nodes_map. 2014-05-14 09:10:03 +02:00
antirez cdf2271c5b cluster.tcl: saner error handling.
Better handling of connection errors in order to update the table and
recovery, populate the startup nodes table after fetching the list of
nodes.

More work to do about it, it is still not as reliable as
redis-rb-cluster implementation which is the minimal reference
implementation for Redis Cluster clients.
2014-05-14 00:15:52 +02:00
antirez bae30479fb redis.tcl: return I/O error message when peer closes connection. 2014-05-14 00:14:35 +02:00
antirez 76c31d425e Scripting test: check that Lua can call commands rewirting argv.
SPOP, tested in the new test, is among the commands rewritng the
client->argv argument vector (it gets rewritten as SREM) for command
replication purposes.

Because of recent optimizations to client->argv caching in the context
of the Lua internal Redis client, it is important to test for SPOP to be
callable from Lua without bad effects to the other commands.
2014-05-07 16:12:32 +02:00
antirez 088b9eadc4 Test: handle new osx 'leaks' error.
Sometimes the process is still there but no longer in a state that can
be checked (after being killed). This used to happen after a call to
SHUTDOWN NOSAVE in the scripting unit, causing a false positive.
2014-05-07 16:12:32 +02:00
antirez 1c130c6b03 Test: cluster/base, check that we can write/read from cluster. 2014-05-02 16:37:12 +02:00
antirez 3bc119c155 Cluster: Tcl cluster client: handle MOVED/ASK. 2014-05-02 15:35:08 +02:00
antirez fcd2065f8e Cluster: Tcl cluster client: slots-nodes map and close method.
Now the client is able to actually run commands in a Redis Cluster
assuming the slots->nodes map is stable.
2014-05-02 10:56:02 +02:00
antirez 5344357f80 Cluster: Tcl cluster client: build nodes representation. 2014-05-02 10:19:28 +02:00
antirez 8b7e23bdde Cluster: Tcl cluster client: get nodes description. 2014-05-02 09:55:27 +02:00
antirez bc8ea04a7d Cluster: Tcl cluster client key -> hashslot. 2014-04-30 18:55:28 +02:00
antirez e8357d0f85 Cluster test: Tcl cluster library initial skeleton. 2014-04-30 15:47:19 +02:00
antirez 1db45ba58c Cluster test: check for state=ok after slot allocation. 2014-04-30 09:29:03 +02:00
antirez 7b5ce1ffb1 Cluster test: slots allocation. 2014-04-29 18:40:43 +02:00
antirez 4a3db25504 Cluster test: use 20 instances.
This makes tests a bit slower, but it is better to test things at a
decent scale instead of using just a few nodes, and for a few tests we
actually need so many nodes.
2014-04-29 16:20:43 +02:00
antirez e8631a6991 Cluster / Sentinel test: instances count moved to run.tcl. 2014-04-29 16:17:15 +02:00
antirez 9e422f74a6 Cluster test: config epoch conflict resolution. 2014-04-29 15:39:59 +02:00
antirez 2c55622333 Cluster test: auto-discovery to form full mesh. 2014-04-29 15:00:11 +02:00
antirez 2555b2f4bd Cluster test: check that every node has a different ID. 2014-04-29 10:42:32 +02:00
antirez e1b129811a Cluster test: basic cluster nodes info access functions. 2014-04-29 10:42:17 +02:00
antirez 93e7a130fc Test: fixed scripting.tcl test false positive. 2014-04-24 21:44:32 +02:00
antirez c3f85c0107 Redis Cluster test framework skeleton. 2014-04-24 18:01:41 +02:00
antirez 897adc1c8c Sentinel test files / directories layout improved.
The test now runs in a self-contained directory.
The general abstractions to run the tests in an environment where
mutliple instances are executed at the same time was extrapolated into
instances.tcl, that will be reused to test Redis Cluster.
2014-04-24 11:08:22 +02:00
antirez e6b6813656 Test: spawn_instance now supports additional config. 2014-04-24 10:50:51 +02:00
Salvatore Sanfilippo 32c917964e Merge pull request #1677 from mattsta/expire-before-delete
Check key expiration before deleting
2014-04-23 16:13:49 +02:00
yoav fdaab02347 Merge remote-tracking branch 'upstream/unstable' into unstable 2014-04-22 10:01:21 +03:00
antirez 9caa1ae96a Fuzzy test for ZREMRANGEBYLEX added. 2014-04-18 13:02:16 +02:00
antirez 85a2f2354e PFCOUNT multi-key test added. 2014-04-18 12:36:33 +02:00
antirez c0ccd4da7d Sorted set lex ranges stress tester. 2014-04-17 10:25:58 +02:00
antirez 5c48432340 Basic ZRANGEBYLEX / ZLEXCOUNT tests. 2014-04-17 00:08:11 +02:00
antirez cffeafe391 More HyperLogLog tests. 2014-04-16 09:17:38 +02:00
antirez 81ceef7d22 PFMERGE fixed to work with sparse encoding. 2014-04-14 16:09:32 +02:00
Matt Stancliff 88530804e8 Add test for deleting an expired key
Verify proper expire-before-delete behavior.

This test passes with the expire-before-delete commit and fails
without it.
2014-04-10 18:32:19 -04:00
yoav 4930d903fc Fix eval usage in tests to conform with eval semantics 2014-04-06 17:20:01 +03:00
antirez 352208ff68 Initial HyperLogLog tests. 2014-04-03 22:16:05 +02:00
antirez 9e0b9f12b2 Test: do not complain when "leaks" can't run because process died. 2014-03-25 09:33:37 +01:00
antirez 258d377d8e Sentinel test: 02 unit better coverage + refactoring. 2014-03-18 15:18:51 +01:00
antirez 58f104e2ee Sentinel test: foreach_instance_id implements 'break'. 2014-03-18 15:06:52 +01:00
antirez 2586ea7637 Sentinel: instance_is_killed proc added to sentinel.tcl. 2014-03-18 14:58:27 +01:00
antirez 8d011492a0 Sentinel test: set less time sensitive defaults.
This commit sets the failover timeout to 30 seconds instead of the 180
seconds default, and allows to reconfigure multiple slaves at the same
time.

This makes tests less sensible to timing, with the result that there are
less false positives due to normal behaviors that require time to
succeed or to be retried.

However the long term solution is probably some way in order to detect
when a test failed because of timing issues (for example split brain
during leader election) and retry it.
2014-03-04 17:10:29 +01:00
antirez efb092baa6 Sentinel test: debugging console improved. 2014-03-04 15:55:36 +01:00
antirez 7d97a4c99b Sentinel test: initial debugging console. 2014-03-04 12:05:49 +01:00
antirez 3072a1e781 Sentinel test: be more patient in create_redis_master_slave_cluster. 2014-03-04 11:20:53 +01:00
antirez dd8d883c9c Sentiel test: add test start time in output. 2014-03-04 11:17:27 +01:00
antirez 35e8bc305d Sentinel test: use 1000 as retry in initial 00 unit test. 2014-03-03 13:26:18 +01:00
antirez 1734f7aedb Sentinel test: initial tests in 03 unit. 2014-03-03 13:25:14 +01:00
antirez db9e718c8a Sentinel test: foreach_instance_id now supports 'continue'. 2014-03-03 13:23:32 +01:00
antirez 69cc3b692f Sentienl test: fixed typo in unit 03 top comment. 2014-03-03 13:01:11 +01:00
antirez 8b8c1cd4c2 BITPOS fuzzy testing. 2014-02-27 15:27:05 +01:00
antirez b21f4d63de Basic BITPOS tests. 2014-02-27 15:01:45 +01:00
antirez 8d95a47408 Added two more BITCOUNT tests stressing misaligned access. 2014-02-27 10:07:29 +01:00
antirez 2a7847a3b5 BITCOUNT fuzzy test with random start/end added.
It was verified in practice that this test is able to stress much more
the implementation by introducing errors that were only trivially to
detect with different offsets but impossible to detect starting always
at zero and counting bits the full length of the string.
2014-02-27 10:00:17 +01:00
antirez 55c059e270 Sentinel test: add stub for unit 04. 2014-02-25 15:36:51 +01:00
antirez fe9a489b29 Sentinel test: added TODO items in 02 unit. 2014-02-25 15:21:53 +01:00
antirez 339322fe58 Sentinel test: check role at end of unit 01. 2014-02-25 14:59:36 +01:00
antirez fef88681f2 Sentinel test: kill masters instead of using DEBUG SLEEP in all tests. 2014-02-25 14:33:44 +01:00
antirez 386467acfb Sentinel test: restart instances left killed by previous unit.
An unit can abort in the middle for an error. The next unit should not
assume that the instances are in a clean state, and must restart what
was left killed.
2014-02-25 08:48:46 +01:00
antirez a9360c62e8 Sentinel test: jump to next unit on test failure.
Sentinel tests are designed to be dependent on the previous tests in the
same unit, so usually we can't continue with the next test in the same
unit if a previous test failed.
2014-02-25 08:33:41 +01:00
antirez 044b627549 Sentinel test: test majority crashing Sentinels.
The test was previously performed by removing the master from the
Sentinel monitored masters. The test with the Sentinels crashed is
more similar to real-world partitions / failures.
2014-02-25 08:29:12 +01:00
antirez 630fb3539f Sentinel test: restart_instance should refresh pid attrib.
Also kill_instance was modified to warn when a test will try to kill the
same instance multiple times for error.
2014-02-25 08:23:48 +01:00
antirez d3a3ef0bc1 Sentinel test: more stuff mored from 00-base to init.
The area a number of mandatory tests to craete a stable setup for
testing that is not too sensitive to timing issues. All those tests
moved to includes/init-tests, and marked as (init).
2014-02-24 17:21:50 +01:00
antirez 29f4df8018 Sentinel test: removed useless code to set SDOWN timeout.
The new common initialization code used to start a new unit already set
the timeout to 2000 milliseconds.
2014-02-24 16:57:52 +01:00
antirez 540536c055 Sentinel test: tmp dir and gitignore added. 2014-02-24 11:51:31 +01:00
antirez 09dec3613e Sentinel test: minor fixes to --pause-on-error. 2014-02-23 18:02:52 +01:00
antirez afd3db17a0 Sentinel test: --pause-on-error option added.
Pause the test with running instances available for state inspection on
error.
2014-02-23 17:57:56 +01:00
antirez a929867cca Sentinel test: added empty units to fill later. 2014-02-23 17:50:59 +01:00
antirez 8c254415f7 Sentinel test: framework improved and conf-update unit added.
It is now possible to kill and restart sentinel or redis instances for
more real-world testing.

The 01 unit tests the capability of Sentinel to update the configuration
of Sentinels rejoining the cluster, however the test is pretty trivial
and more tests should be added.
2014-02-22 17:27:49 +01:00
antirez d7da507683 Sentinel test: move init tests as includes.
Most units will start with these two basic tests to create an
environment where the real tests are ran.
2014-02-20 16:58:23 +01:00
antirez 5765444454 Sentinel test: ability to run just a subset of test files. 2014-02-20 16:28:41 +01:00
antirez e087d8a20d Sentinel test: some reliability fixes to 00-base tests. 2014-02-19 10:26:23 +01:00
antirez a88a057a1f Sentinel test: check that role matches at end of 00-base. 2014-02-19 10:08:49 +01:00