4c696d5bf2
Merge branch 'release-6.2' into dd-better-rebalance-logging
...
# Conflicts:
# fdbserver/DataDistributionQueue.actor.cpp
2020-02-21 17:41:00 -08:00
dfa5f76c01
Remove unused parameter. Don't put check for g_network presence in ASSERT_WE_THINK.
2020-02-21 16:28:03 -08:00
08914a2acd
Once available space ratio falls below 0.3 avoid moving data to teams with less free space than the median team
2020-02-21 15:14:32 -08:00
2e699fef55
Don't suppress actor cancellation because we've already initialized the trace event by adding details.
2020-02-21 11:28:59 -08:00
6810a03283
Add more logging to valley filler and mountain chopper
2020-02-21 10:55:14 -08:00
819c55556c
More aggressively attempt to find teams that do not have low disk space
2020-02-20 16:47:50 -08:00
e1fb568fd1
Merge branch 'release-6.2' into dd-use-available-space
...
# Conflicts:
# fdbserver/DataDistribution.actor.cpp
# fdbserver/DataDistribution.actor.h
# fdbserver/DataDistributionQueue.actor.cpp
2020-02-20 16:12:42 -08:00
e4b483796d
Combine some logic that was doing similar computations for free space ratio.
2020-02-20 14:52:08 -08:00
4c9c736253
Data distribution uses available space instead of free space when evaluating whether processes are low on space and penalizing them.
2020-02-20 11:21:03 -08:00
c164acb88d
Add new criteria to DD's GetTeamRequest that allow you to require shards be present on the team and that the team have a minimum free ratio. This avoids scenarios where the team chosen when processing the request is later rejected by the requestor, causing rebalancing movements to get stuck.
2020-02-20 09:32:00 -08:00
abb75f7eb7
Add logging to indicate the time spent at each priority that exceeds some minimum busyness threshold
2020-02-07 14:34:24 -08:00
3f9d9d8b84
Merge branch 'release-6.2'
...
# Conflicts:
# CMakeLists.txt
# cmake/FlowCommands.cmake
# documentation/sphinx/source/release-notes.rst
# fdbclient/StorageServerInterface.h
# fdbserver/DataDistributionTracker.actor.cpp
# fdbserver/MasterProxyServer.actor.cpp
# fdbserver/fdbserver.actor.cpp
# flow/Knobs.h
# flow/Platform.cpp
# versions.target
2020-01-16 18:37:47 -08:00
e65760eb46
Merge pull request #2536 from etschannen/feature-commit-latency
...
Improved commit latency in large clusters
2020-01-13 19:12:02 -08:00
ab7071932f
Data distribution no longer attempts to pick teams which share members of the source unless the team matches exactly
2020-01-09 16:59:37 -08:00
3157d8a375
fixed typo
2019-12-18 16:57:39 -08:00
688940b685
merge 6.2 into master
2019-10-21 11:43:46 -07:00
5667331729
added a buggify + minor code cleanup
2019-10-11 18:31:43 -07:00
86bcb84b45
Raised the data distribution priority of splitting shards above restoring fault tolerance to avoid hot write shards
2019-10-11 17:50:43 -07:00
22c3fa867c
Merge pull request #2074 from xumengpanda/mengxu/fix-correctness-bug
...
DD:fix:getTeam may fail to get a team when it can
2019-10-07 09:33:57 -07:00
3c4dc1003d
DD:clang-format the PR
2019-09-11 11:16:29 -07:00
a83afa5f64
DD:BgDDValleyFiller:Do not move shard to a team with no healthy space
2019-09-11 11:15:00 -07:00
bd80a67d46
Merge branch 'master' into mengxu/storage-engine-switch-PR-v2
2019-09-03 14:11:33 -07:00
a377261740
StorageEngineSwitch:Remove questions in comments
2019-08-22 11:49:39 -07:00
ac68c8e4fd
added sources servers to the warning message
2019-08-21 11:48:29 -07:00
39680fa515
StorageEngineSwitch:Clean up unnecessary trace
...
And do not trigger storage recruitment unnecessarily.
2019-08-19 14:11:57 -07:00
d30d4cb955
Added a duration to regular relocateShard trace events
2019-08-16 15:15:36 -07:00
297b65236f
added additional trace events to warn when different parts of shard relocations take more than 10 minutes
2019-08-16 14:56:58 -07:00
a588710376
StorageEngineSwitch:Graceful switch
...
When fdbcli change storeType for storage engines,
we switch the store type of storage servers one by one gracefully.
This avoids recruiting multiple storage servers on the same process,
which can cause OOM error.
2019-08-12 17:37:52 -07:00
b653ddb30d
Final clean ups after rebasing master
2019-07-30 22:35:34 -07:00
cda70700cc
Address review comments. 50K correctness with no failures.
2019-07-30 22:24:30 -07:00
5d20364423
Address review comments
2019-07-30 22:24:30 -07:00
1922c39377
Resolve review comments. 100K run shows one suspecious ASSERT_WE_THINK failure which I think could be a race.
2019-07-30 22:24:30 -07:00
f5d6e3a5b3
- Addressed review commends
...
- Added test for the storage server failure disable switch
2019-07-30 22:20:45 -07:00
4ecfc9830f
Added finer grained controls to DataDistribution in fdbcli. What's happening under the hood is:
...
- Use pre-existing 'healthZone' key and write a special value to it in order to disable DD for all storage server failures
- Use a new system key 'rebalanceDDIgnored' key to disable/enable DD for all rebalance reasons(MountainChopper and ValleyFiller)
Kicked off two 200K correctness and showed no related errors.
2019-07-30 22:17:21 -07:00
a78a97f186
Merge pull request #1908 from etschannen/feature-better-dd
...
A few data distribution improvements
2019-07-30 17:34:50 -07:00
63941e0d96
disable DD with a in-memory flag and use in snapv2
2019-07-30 17:04:51 -07:00
5dd9043fd3
addressed review comments
2019-07-30 17:04:41 -07:00
481642fbd4
Merge branch 'master' into feature-better-dd
2019-07-30 16:56:27 -07:00
41605735f5
Merge pull request #1916 from ajbeamon/merge-onto-new-servers
...
Add knob to control whether merges request new servers or not.
2019-07-30 15:04:37 -07:00
14648e20f9
Merge pull request #1901 from ajbeamon/data-distribution-receives-bytes-input-rate
...
Send bytes input rate to data distribution
2019-07-30 15:01:36 -07:00
bc536757df
Add knob to control whether merges request new servers or not. Set the default to request new servers in \xff but not in main key space.
2019-07-29 15:47:34 -07:00
6b5e683de5
The mountainChopper and valleyFiller only move larger than average shards, to avoid moving high bandwidth shards which are generally smaller.
2019-07-28 23:50:42 -07:00
04dd293af0
Merge pull request #1874 from xumengpanda/mengxu/DD-code-read
...
DataDistribution:Add comments to help understand the code
2019-07-26 13:30:44 -07:00
b91795d288
Send bytes input rate to DD.
2019-07-25 16:27:32 -07:00
e582219ec5
Remove unnecessary condition in DDQueue
...
Resolve the review comment.
2019-07-22 17:00:37 -07:00
b7478f5dd3
DD:Add comments to help understand code
...
Add comments to explain the functionalities of some code.
2019-07-22 11:23:16 -07:00
612a51fe00
Apply Clang format to PRIORITY_TEAM_REDUNDANT
2019-07-19 18:32:22 -07:00
ea76451f15
Count PRIORITY_TEAM_REDUNDANT as count PRIORITY_TEAM_UNHEALTHY
2019-07-19 18:30:01 -07:00
7a500cd37f
A giant translation of TaskFooPriority -> TaskPriority::Foo
...
This is so that APIs that take priorities don't take ints, which are
common and easy to accidentally pass the wrong thing.
2019-06-25 02:47:35 -07:00
5f55f3f613
Replace g_random and g_nondeterministic_random with functions deterministicRandom() and nondeterministicRandom() that return thread_local random number generators. Delete g_debug_random and trace_random. Allow only deterministicRandom() to be seeded, and require it to be seeded from each thread on which it is used.
2019-05-10 14:01:52 -07:00
2d5043c665
Merge branch 'release-6.1'
...
# Conflicts:
# documentation/sphinx/source/release-notes.rst
# versions.target
2019-04-30 18:27:04 -07:00
e0f7ec96aa
Data distribution needs to build new teams as old teams are removed to ensure data remains balanced across servers
2019-04-22 17:29:46 -07:00
d01cbf3455
Addressed code review comments
2019-04-05 13:12:20 -07:00
1c16f87a4e
Remove trace-calls to printable (in non-workloads)
2019-04-05 13:12:19 -07:00
981426bac9
More ide fixes
2019-03-05 18:03:57 -08:00
d008de576e
Merge pull request #1139 from xumengpanda/mengxu/machine-team-upgrade-PR
...
Add background actor to remove redundant teams
2019-02-22 14:22:07 -08:00
9445ac0b0c
Status: Use new data distributor worker to publish status
...
After we add a new data distributor role, we publish the data
related to data distributor and rate keeper through the new
role (and new worker).
So the status needs to contact the data distributor, instead of master,
to get the status information.
2019-02-21 18:05:50 -08:00
7cca439e00
TeamRemover: Add status to show redundant team removing
...
Distinguish the removal of unhealthy team and redundant team.
Change status report to include redundant team removal report.
2019-02-21 14:16:46 -08:00
27a3153719
Use ACTOR forward declarations in MoveKeys
...
Also MoveKeys.h -> MoveKeys.actor.h
2019-02-19 15:16:59 -08:00
3a0f9839b9
Fix minor IDE build errors
2019-02-19 15:16:59 -08:00
6d09ac483c
Merge with master
2019-02-15 17:03:40 -08:00
bf6da81bf9
Remove recovery version from data distribution queue
...
This parameter is no longer used/needed.
2019-02-14 16:37:16 -08:00
07dab56133
Fix a data movement stuck bug
...
When moving keys to a team, if one of the server in the target team died, then
the move can become stuck. This is because the DDTeamCollection waits for all
the data movement of the failed server to be completed. However, in this case,
because the movement has not finished yet, checking the database tells us there
is no key assocated with this server and it is safe to go ahead. In reality,
only the in-memory structure knows there is pending movement, i.e., unfinished
move causes some keys to be attributed to the failed server. Thus, the server
can't be removed yet. Fix by adding a check with in-memory structure in
waitForAllDataRemoved().
Use const& to optimize a few function parameters.
2019-02-14 16:37:16 -08:00
886e7ab2ba
Add a new DataDistributor role.
...
Let cluster controller to start a new data distributor role by sending a
message to a chosen worker.
Change MasterInterface usage in DataDistribution to masterId
Add DataDistributor rejoin handling.
This allows the data distributor to tell the new cluster controller of its
existence so that the controller doesn't spawn a new one. I.e., there should
be only ONE data distributor in the cluster.
If DataDistributor (DD) doesn't join in a while, then ClusterController (CC) tries
to recruit one as DD. CC also monitors DD and restarts one if it failed.
The Proxy is also monitoring the DD. If DD failed, the Proxy will ask CC for
the new DD.
Add GetRecoveryInfo RPC to master server, which is called by data distributor
to obtain the recovery Transaction version from the master server.
2019-02-14 16:30:13 -08:00
fe4f43203d
TeamCollection: getTeam may add a new team
...
getTeam function may add a new team for the GetTeamRequest.
We need to check if the number of teams is larger than the desired team number.
2019-02-12 14:57:35 -08:00
8de031f9a6
TeamCollection: clang-format
...
Format the changes with git clang-format.
No functional changes.
Signed-off-by: Meng Xu <meng_xu@apple.com>
2018-11-21 11:18:26 -08:00
f7a7e069f0
TeamCollection: Remove unnecessary comments
...
Pass 41806 tests with no failure
Signed-off-by: Meng Xu <meng_xu@apple.com>
2018-11-16 15:56:35 -08:00
73c58852f0
TeamCollection: Resolve code review comments
...
Resolve code review comments:
1) Improve the code efficiency by avoiding unnecessary map search
and avoiding unnecessary checking
2) Remove or comment out trace events when they can be spammy
3) Improve coding style
Tested for 1 hour and no error was found.
KillRegionCycle.txt test was excluded from the test because
existing code cannot pass that test either
Signed-off-by: Meng Xu <meng_xu@apple.com>
2018-11-16 15:55:33 -08:00
5051b35c61
TeamCollection: Use machine team to create server team
...
Current server team collection logic does not consider
the fact that multipe storage servers can run on the same machine.
When multiple machines fail, all servers on the machines will fail, and
the possibility of having one process team fail and lose data is very high.
To reduce the possibility of losing data when multiple machine fails,
we first create machine teams which span across different fault zones;
we then create server teams based on machine teams by
first picking 1 machine team, and then
picking 1 server from each machine in the machine team.
Signed-off-by: Meng Xu <meng_xu@apple.com>
2018-11-16 15:53:22 -08:00
4e54690005
Merge branch 'release-6.0'
...
# Conflicts:
# fdbserver/DataDistribution.actor.cpp
# fdbserver/MoveKeys.actor.cpp
2018-11-12 20:26:58 -08:00
cd188a351e
fix: if a destination team became unhealthy and then healthy again, it would lower the priority of a move even though the source servers we are moving from are still unhealthy
...
fix: badTeams were not accounted for when checking priorities
2018-11-11 12:33:31 -08:00
268093a96d
Adjust all includes to be relative to the root.
...
Remove the use of relative paths. A header at foo/bar.h could be included by
files under foo/ with "bar.h", but would be included everywhere else as
"foo/bar.h". Adjust so that every include references such a header with the
latter form.
Signed-off-by: Robert Escriva <rescriva@dropbox.com>
2018-10-19 17:35:33 +00:00
90301f497f
Merge branch 'release-6.0'
...
# Conflicts:
# fdbclient/ManagementAPI.actor.cpp
# fdbrpc/FlowTransport.actor.cpp
# fdbrpc/TLSConnection.actor.cpp
# fdbserver/DataDistribution.actor.cpp
# fdbserver/Status.actor.cpp
# fdbserver/storageserver.actor.cpp
# fdbserver/workloads/StatusWorkload.actor.cpp
# versions.target
2018-09-05 16:06:33 -07:00
d8659a5822
fix: bytesWritten would overflow and go negative
2018-08-31 12:46:57 -07:00
2a97139d5d
This is the first step in eliminating the usage of database names in our code. The C API remains the same, but underneath that all usage of database names is eliminated.
2018-08-16 10:24:12 -07:00
fb31a6999f
Rewrite all files to have #include actorcompiler.h as the last include.
2018-08-14 15:50:26 -07:00
535b5701e5
Rewrite all `Void _ = wait(...)` -> `wait(...)`.
...
This takes advantage of the new actorcompiler functionality to avoid
having duplicate definitions of `Void _` when trying to feed the
un-actorompiled source through clang.
2018-08-14 15:50:26 -07:00
574c5576a2
Merge branch 'release-6.0' of github.com:apple/foundationdb
...
# Conflicts:
# fdbrpc/TLSConnection.actor.cpp
# versions.target
2018-08-10 14:31:58 -07:00
6f02ea843a
prevented a slow task when too many shards were sent to the data distribution queue after switching to a fearless deployment
2018-08-09 12:37:46 -07:00
3535ddad80
Merge pull request #674 from alexmiller-apple/glibcxx-debug-fixes
...
Fix bugs uncovered by -D_GLIBCXX_DEBUG
2018-08-09 08:18:51 -07:00
1a7cda4149
Stop performing self-moves. (e.g. a = std::move(a))
...
self-moves are frowned upon in C++, and in our code this generally happens from
calls to swap as part of trying to implement a "unordered erase" function via
swap-to-the-end-and-pop_back. For convenience, a swapAndPop() function is now
offered that performs this, while disallowing self-moves.
2018-08-01 18:09:54 -07:00
1c29275672
call all methods which could disable a trace event before it is initialized. In practice this means calling .error first, then .suppressFor, then all your details.
2018-08-01 14:30:57 -07:00
d6c6e7d306
fix: do not attempt data movement to an unhealthy destination team
...
allow building more teams than desired if all teams are unhealthy
bestTeamStuck is an error in simulation again
2018-07-07 16:51:16 -07:00
e951df95b7
when doing data movement where one region has the data and the other doesn’t, first move a single replica to the other region to save WAN bandwidth
2018-06-19 23:15:30 -07:00
1ccfb3a0f4
fix: log_anti_quorum was always 0 in simulation
...
removed durableStorageQuorum, because it is no longer a useful configuration parameter
2018-06-18 10:24:57 -07:00
372ed67497
Merge branch 'master' into feature-remote-logs
...
# Conflicts:
# fdbserver/DataDistribution.actor.cpp
# fdbserver/MasterProxyServer.actor.cpp
# fdbserver/TLogServer.actor.cpp
# fdbserver/TagPartitionedLogSystem.actor.cpp
2018-06-11 11:34:10 -07:00
b9826dc1cb
fix: do not automatically reduce redundancy we move keys if the database does not have remote replicas. This is to prevent problems when dropping remote replicas from a configuration.
2018-06-08 16:17:27 -07:00
e5488419cc
Attempt to normalize trace events:
...
* Detail names now all start with an uppercase character and contain no underscores. Ideally these should be head-first camel case, though that was harder to check.
* Type names have the same rules, except they allow one underscore (to support a usage pattern Context_Type). The first character after the underscore is also uppercase.
* Use seconds instead of milliseconds in details.
Added a check when events are logged in simulation that logs a message to stderr if the first two rules above aren't followed.
This probably doesn't address every instance of the above problems, but all of the events I was able to hit in simulation pass the check.
2018-06-08 11:11:08 -07:00
7d392689fe
fix: only update metrics for healthy destinations, because unhealthy destinations are already in the source
2018-06-07 18:12:04 -07:00
b423d73b42
fix: do not finish a shard relocation until all of the storage servers have made the current recovery version durable. This is to prevent dropping a needed storage server as a source for a shard after dropping a remote configuration
2018-06-07 12:29:25 -07:00
10d25927cd
Merge branch 'master' into feature-remote-logs
...
# Conflicts:
# fdbserver/DataDistribution.actor.cpp
2018-04-30 22:15:39 -07:00
19762b847d
Merge branch 'release-5.2'
...
# Conflicts:
# fdbserver/DatabaseConfiguration.cpp
# fdbserver/SimulatedCluster.actor.cpp
2018-04-10 17:02:43 -07:00
7af892f50b
first working version of non-copying recovery working with fearless configurations
2018-04-08 21:24:05 -07:00
da737e1ea3
suppress the BestTeamStuck trace event
2018-03-26 18:32:32 -07:00
3abf4d7fdf
Merge branch 'master' into feature-remote-logs
2018-03-09 14:50:04 -08:00
91bb8faa45
Merge commit 'f773b9460d31d31b7d421860fc647936f31aa1fa'
...
# Conflicts:
# tests/fast/SidebandWithStatus.txt
# tests/rare/LargeApiCorrectnessStatus.txt
# tests/slow/DDBalanceAndRemoveStatus.txt
2018-03-09 14:47:03 -08:00
cf6dd1437b
suppress spammy trace events
2018-03-09 10:16:34 -08:00
5390af8be4
suppress spammy logs
2018-03-09 09:40:36 -08:00
fa7eaea7cf
fix: shards affected by team failure did not properly handle separate teams for the remote and primary data centers
2018-03-08 10:50:05 -08:00
37a6a81634
Merge commit '7f6fc3e039c911cd84b8540f7f799fc38a1c1822' into feature-remote-logs
...
# Conflicts:
# fdbserver/workloads/RestartRecovery.actor.cpp
2018-02-23 12:33:28 -08:00
A.J. Beamon