23706c957b
Use DDSketch for Sample Data.
2022-11-12 13:45:46 -08:00
49f0cf5ab0
Force name and description of workloads to be the same
2022-10-13 20:53:48 -06:00
4fd64630e8
Convert literal string ref instances to use _sr suffix
2022-09-19 11:35:58 -07:00
0c2bd73e28
Fix typo in SHARD_READ_HOT_BANDWIDTH_MIN_PER_KSECONDS name
2022-06-27 09:56:02 -07:00
a71099471b
Update copyright header dates
2022-03-21 13:36:23 -07:00
138932c12f
Fixed a memory bug in the watch map. A few other formatting tweaks, etc.
2022-01-05 11:26:43 -08:00
d8e161f89e
Refactor NativeAPI transactions to create and pass around a reference counted state object. Watches no longer use the tranasction info object but instead use their own state.
2021-12-17 11:57:39 -08:00
abf73047ca
Enforce std:: specifier rather than using namespace
2021-09-16 19:40:28 -07:00
df90cc89de
apply clang-format to *.c, *.cpp, *.h, *.hpp files
2021-03-10 10:18:07 -08:00
7f0d14c8e4
Modernize/refactor workloads directory
2020-10-04 22:29:07 -07:00
974c9248f8
The workload sometimes would wrongly report a passed tests as failed due to not return after the hotkey is found.
2020-08-31 14:09:01 -07:00
e13e7c5725
Update fdbserver/workloads/ReadHotDetection.actor.cpp
...
Co-authored-by: Jingyu Zhou <jingyuzhou@gmail.com>
2020-08-12 15:48:01 -07:00
3ac6996844
Increased some knobs to throttle the spammy read hot logging. Also added more details inside the read hot log to make it useful
2020-08-12 14:27:34 -07:00
bdc58eaa11
Fix invalid memory access in ReadHotDetection test due to memory being stored in a non-standalone stringref.
2020-05-12 11:16:35 -07:00
dd9c240320
Address the remaining review comments
2020-05-04 10:25:10 -07:00
1927f15898
Now the ReadHotDetection tests can be run reliably. Added it to nightly bundle.
2020-04-28 13:26:02 -07:00
b0a1af1288
Added the actual read hot detection algorithm and logging machanism.
...
- When a shard has a read bandwidth larger than a threshold value(configurable via knob), and it's read-bandwidth/byte-size ratio is also larger than a threshold(configurable via knob), the corresponding shard tracker will run the algorithm
- The algorithm will divide the shard into 10MB(configurable via knob) chunks and try to find the chunk(s) that has large aforementioned ratio
- Then those ranges will be logged into TraceEvents. This will later do more like actually cache them.
2020-01-21 11:19:52 -08:00
Sam Gwydir