afd3ec13ff
added knobs
2020-01-21 18:58:34 -08:00
7a4b459f07
wait for a tls handshake to complete before returning a connection
...
wait for multiple tls errors before throttling
2020-01-21 16:45:15 -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
0e916fdbed
throttle client TLS errors longer than server errors so that when both happen simultaneously the server throttling will be disabled when the client makes its next attempt
2020-01-12 22:12:18 -08:00
1f7eb1f738
throttle outgoing tls connections before establishing a network connection
...
store serverTLSConnectionThrottler map inside of g_network, so that it works properly with simulation
2020-01-12 16:44:30 -08:00
ef5dfb87dc
Merge pull request #2529 from bnamasivayam/tls-throtlling
...
Establishing TLS connection through the handshake process is expensiv…
2020-01-12 14:56:21 -08:00
741aa523e6
Establishing TLS connection through the handshake process is expensive and the fdbserver process can get easily saturated with doing repeated TLS handshakes with only a few hundreds of clients have bad certificate. Hence throttle the number of handshakes done on the server per client ip if it has a bad certificate.
2020-01-10 16:19:41 -08:00
2e20c12200
Merge pull request #2475 from ajbeamon/priority-busy-fixes
...
Fix PriorityBusy calculation and add PriorityMaxBusy
2020-01-10 12:47:17 -08:00
176a1b6319
Merge pull request #2515 from ajbeamon/remove-timer-in-slowtask-profiler
...
Fix slow task profiler crash
2020-01-10 12:41:57 -08:00
a5f544818c
Merge pull request #2420 from ajbeamon/trace-clock-source-fix
...
Revert change to make g_trace_clock thread_local, ...
2020-01-10 12:36:38 -08:00
de5a591b15
Attempt a minor pointless change to fix the build
2020-01-06 15:17:13 -08:00
6cf38790d6
Reorganize declaration of variable and add release note.
2020-01-06 12:27:56 -08:00
4a52864023
Remove call of timer() from the slow task profiling signal handler, as it can lead to crashes if called at the wrong time.
2020-01-06 12:19:45 -08:00
16b5af067c
changed trace event name
2020-01-03 16:03:29 -08:00
deb032745a
fix: do not set logged until then end of the function
2020-01-03 12:45:23 -08:00
1867d30017
added asserts to protect against future actions on a trace event that has been logged
2020-01-03 12:31:06 -08:00
7152469cc3
log the base trace event before the endpoint messages
2020-01-03 12:15:38 -08:00
6e473c3a83
Merge branch 'release-6.2' into feature-addpeer-fix
2020-01-02 17:37:23 -08:00
032797ca5c
Merge pull request #2430 from etschannen/release-6.2
...
Reduce recovery times caused by saturating the cluster controller
2020-01-02 17:35:59 -08:00
3dd3ac3cfd
Merge branch 'release-6.2' into trace-clock-source-fix
...
# Conflicts:
# documentation/sphinx/source/release-notes.rst
2020-01-02 15:14:12 -08:00
ca01593067
Cap busyness to 1.0 at logging time to cover all cases where it could be measured above.
2020-01-02 15:10:42 -08:00
9e137d3b49
fix: addPeerReference only marks a connection as healthy if it is the first peerReference
...
added additional logging to long LoadBalance calls, and when the failure monitor state changes for an address
2019-12-19 18:26:29 -08:00
3b28c7f103
Throw the correct error in deleteFile
2019-12-19 14:13:09 -08:00
d8c3c2fda4
Improved prioritization of commit path on the proxies
2019-12-18 16:56:35 -08:00
a093021855
Fix priority time calculation. Track max priority busy rather than seconds squared.
2019-12-17 09:14:54 -08:00
5080b8293a
Added Windows required library for function GetProcessMemoryInfo
2019-12-16 08:09:10 -08:00
3c30215662
Merge branch 'release-6.2' of github.com:apple/foundationdb into release-6.2
2019-12-09 13:18:07 -08:00
20eacdb434
Add missing include
2019-12-06 15:18:17 -08:00
9866d1ce27
Revert change to make g_trace_clock thread_local, instead checking we are on the correct thread when getting the time.
2019-12-06 10:15:49 -08:00
78b202f3a4
Apply A.J.'s suggestion to randomInt as well
2019-12-05 11:01:41 -08:00
cf5cdc4e93
Update flow/DeterministicRandom.cpp
...
Include equality now that we've adjusted the value by 1.
Co-Authored-By: A.J. Beamon <ajbeamon@users.noreply.github.com>
2019-12-05 11:01:03 -08:00
b09f0b334b
Take A.J.'s suggestion, which fixes A.J.s counterexample
2019-12-05 10:27:35 -08:00
96e71bb109
Fix old Make build
2019-12-04 17:06:49 -08:00
89a093e035
Accept UBSAN's suggestion
...
/home/anoyes/workspace/foundationdb/flow/DeterministicRandom.cpp:72:29: runtime error: negation of -9223372036854775808 cannot be represented in type 'long int'; cast to an unsigned type to negate this value to itself
2019-12-04 16:45:45 -08:00
4f943be21d
Move DeterministicRandom impl to its own translation unit
...
This will allow me to recompile faster after making changes, and should
(slightly) speed up overall compilation.
I manually verified that the unseed matched for one test before and
after this change, so I probably didn't screw up the refactor
2019-12-04 16:45:32 -08:00
5a6bc2aa71
increase the priority of cluster controller recruitment to prefer recruitment over sending serverDBInfo
2019-12-04 16:28:41 -08:00
5f1ef53f62
increase the priority at which the cluster controller registers workers to avoid having a saturated cluster controller recruit a master without all available workers
2019-12-04 16:17:41 -08:00
46d10dc7dc
Fix "null passed as argument declared not null"
...
Fix several such reports from ubsan
E.g.
/Users/anoyes/workspace/foundationdb/flow/Arena.h:794:16: runtime error: null pointer passed as argument 1, which is declared to never be null
2019-12-03 14:46:53 -08:00
4022ca1381
Add USE_UBSAN cmake option
2019-12-02 16:20:06 -08:00
7f263a2614
Require callers to alignment requirements for aligned_alloc
2019-12-02 15:30:54 -08:00
ff8758b1fd
Request alignment that's at least sizeof(void*)
...
According to https://en.cppreference.com/w/c/memory/aligned_alloc#Notes ,
aligned_alloc may return nullptr if it doesn't like the requested
alignment. Let's also detect if nullptr is returned.
2019-12-02 12:51:33 -08:00
a2c84d932d
Merge pull request #2396 from atn34/atn34/fix-invalid-vptr
...
Fix invalid vptr
2019-11-27 21:16:49 -08:00
7eaf76bacf
Merge pull request #2389 from atn34/atn34/default-init-flatbuffers
...
Default initialize absent flatbuffers members
2019-11-27 21:16:27 -08:00
8fc74e3182
Fix UBSAN error
...
Since QuorumCallback<T> is a non trivial type, we need to construct it
before we interact with it
This change fixes the following UBSAN message
/Users/anoyes/workspace/foundationdb/flow/genericactors.actor.h:930:18: runtime error: member access within address 0x0001243f63d0 which does not point to an object of type 'Callback<Standalone<StringRef> >'
0x0001243f63d0: note: object has invalid vptr
2019-11-27 13:16:48 -08:00
c4e01301b0
Fix a potential UB instance
...
Writing a value which is not 0 or 1 to the underlying memory of a bool
is undefined behavior. Conformant flatbuffers implementations must
accept bytes that are not 0 or 1 as booleans [1]. (Conformant
implementations are only allowed to write the byte 0 or 1 as a boolean
[1])
So this protects us from undefined behavior if we ever read a
flatbuffers message written by an almost-conformant implementation.
[1]: https://github.com/dvidelabs/flatcc/blob/master/doc/binary-format.md#boolean
2019-11-26 11:18:17 -08:00
17ab2f8e00
Default initialize absent flatbuffers members
2019-11-26 10:58:29 -08:00
27cb299d84
simulation can sometimes randomly hang or throw connection_failed, instead of always doing one or the other
2019-11-21 16:24:18 -08:00
2727b91c46
simulation tests network connections failing due to errors instead of just hanging
2019-11-21 12:33:07 -08:00
1e5677b55a
increase the priority of reboot and recruitment requests
2019-11-11 15:17:11 -08:00
325a8e4213
remove confusing USE_ODIRECT knob
2019-10-24 11:44:03 -07:00
Evan Tschannen