Mikhail f. Shiryaev
f728f97350
Add cgroupsv2 for getCGroupLimitedCPUCores
2023-06-28 22:19:44 +02:00
Robert Schulze
33a719b9be
Update getNumberOfPhysicalCPUCores.cpp
2023-01-29 21:37:19 +01:00
Robert Schulze
1bf280cb50
Handle obscure case that /proc/cpuinfo is unmounted
2023-01-29 17:25:08 +00:00
Robert Schulze
094e42d1dc
Fix style
2023-01-24 12:57:56 +00:00
Robert Schulze
36d6af0e04
Yet another build fix
2023-01-24 09:57:12 +00:00
Robert Schulze
1a1fa4a40a
Fix ARM build
2023-01-23 19:48:09 +00:00
Robert Schulze
d07b7ce1ee
Some fixups
2023-01-22 17:25:59 +00:00
Robert Schulze
a733303184
Disable HyperThreading on big machines
2023-01-06 19:59:54 +00:00
Robert Schulze
11fb78191c
(draft) Remove SMP deactivation on big machines
...
This might be a bit controversial but
1. AMD offers 48/67/64 (true) core systems already (Milan, (*)), up to
96 cores/CPU soon (Genoa),
2. Intel sells a few Xeon CPUs with > cores (**),
and we shouldn't artificially limit core utilization on such sytems.
(*) https://en.wikipedia.org/wiki/Epyc
(**) https://www.intel.com/content/www/us/en/products/details/processors/xeon/scalable/platinum/products.html
2023-01-06 10:18:00 +00:00
Nikita Taranov
abd585a38a
Increase threshold for using physical cores ( #42503 )
2022-10-21 12:52:46 +02:00
Robert Schulze
78fc36ca49
Generate config.h into ${CONFIG_INCLUDE_PATH}
...
This makes the target location consistent with other auto-generated
files like config_formats.h, config_core.h, and config_functions.h and
simplifies the build of clickhouse_common.
2022-09-28 12:48:26 +00:00
Alexey Milovidov
fa62c7e982
Fix half of trash
2022-09-10 04:08:16 +02:00
Alexey Milovidov
69b91c5410
Simplification #36313
2022-04-16 02:08:06 +02:00
Alexey Milovidov
364f10aff5
Remove trash
2022-04-15 22:56:51 +02:00
Alexey Milovidov
58fc4e033a
Remove trash
2022-04-15 22:54:08 +02:00
filimonov
24d72a18d0
Update getNumberOfPhysicalCPUCores.cpp
...
In Kubernetes shares are calculated from resources.requests.cpu, quota & period - from limits.
When you configure only requests without limits - the container still can use all the cores. So you see all the cores the system has.
(because of requests the pod just can't be scheduled on the node which have less cpus)
So the effective CPU count is all the available cores if you don't have limits, or calculated from quota & period if you have the limits.
2022-03-31 17:21:54 +02:00
JaySon-Huang
e155057cc5
Fix cpu count with cgroup limits
...
Signed-off-by: JaySon-Huang <tshent@qq.com>
2022-01-21 21:03:23 +08:00
Azat Khuzhin
9926f336e3
Cleanup cpuid contrib
2022-01-20 10:02:03 +03:00
Vitaly Baranov
668653600c
Use SettingMaxThreads only in Settings, call getNumberOfPhysicalCPUCores() instead of SettingMaxThreads::getAuto().
2020-07-31 19:11:27 +03:00
Ivan
85d783c247
Poco contrib refactoring ( #10396 )
...
* Remove config_common.h
* Refactor libcpuid contrib
* Remove support for libcpuinfo
* Define USE_CPUID in Arcadia
* Refactor Poco libraries
2020-05-08 17:11:19 +03:00
Ivan Lezhankin
e230632645
Changes required for auto-sync with Arcadia
2020-04-16 15:31:57 +03:00
Ivan Lezhankin
06446b4f08
dbms/ → src/
2020-04-03 18:14:31 +03:00
