From 279df2ffcc1de5b6f8c584b23998f426257ae983 Mon Sep 17 00:00:00 2001 From: Xunlei Pang Date: Wed, 4 Sep 2019 15:30:02 +0800 Subject: [PATCH 0758/2944] alinux: doc: Add Documentation/alibaba/interfaces.rst This file collects all the interfaces specific to Alibaba Cloud Kernel. Add "memory.wmark_min_adj", "memory.exstat", and "zombie memcgs reaper" descriptions. Reviewed-by: Yang Shi Reviewed-by: Gavin Shan Signed-off-by: Xunlei Pang --- Documentation/alibaba/interfaces.rst | 124 +++++++++++++++++++++++++++++++++++ 1 file changed, 124 insertions(+) create mode 100644 Documentation/alibaba/interfaces.rst diff --git a/Documentation/alibaba/interfaces.rst b/Documentation/alibaba/interfaces.rst new file mode 100644 index 0000000..e874e8d --- /dev/null +++ b/Documentation/alibaba/interfaces.rst @@ -0,0 +1,124 @@ +.. SPDX-License-Identifier: GPL-2.0+ + +============================= +AliOS Cloud kernel interfaces +============================= + +This file collects all the interfaces specific to AliOS Cloud kernel. + +memory.wmark_min_adj +==================== + This file is under memory cgroup v1. + + In co-location environment, there are more or less some memory + overcommitment, then BATCH tasks may break the shared global min + watermark resulting in all types of applications falling into + the direct reclaim slow path hurting the RT of LS tasks. + (NOTE: BATCH tasks tolerate big latency spike even in seconds + as long as doesn't hurt its overal throughput. While LS tasks + are very Latency-Sensitive, they may time out or fail in case + of sudden latency spike lasts like hundreds of ms typically.) + + Actually BATCH tasks are not sensitive to memory latency, they + can be assigned a strict min watermark which is different from + that of LS tasks(which can be aissgned a lenient min watermark + accordingly), thus isolating each other in case of global memory + allocation. + + memory.wmark_min_adj stands for memcg global WMARK_MIN adjustment, + it is used to realize separate min watermarks above-mentioned for + memcgs, its valid value is within [-25, 50], specifically: + negative value means to be relative to [0, WMARK_MIN], + positive value means to be relative to [WMARK_MIN, WMARK_LOW]. + + For examples:: + + -25 means memcg WMARK_MIN is "WMARK_MIN + (WMARK_MIN - 0) * (-25%)" + 50 means memcg WMARK_MIN is "WMARK_MIN + (WMARK_LOW - WMARK_MIN) * 50%" + + Negative memory.wmark_min_adj means high QoS requirements, it can + allocate below the global WMARK_MIN, which is kind of like the idea + behind ALLOC_HARDER, see gfp_to_alloc_flags(). + + Positive memory.wmark_min_adj means low QoS requirements, thus when + allocation broke memcg min watermark, it should trigger direct reclaim + traditionally, and we trigger throttle instead to further prevent them + from disturbing others. The throttle time is simply linearly proportional + to the pages consumed below memcg's min watermark. The normal throttle + time once should be within [1ms, 100ms], and the maximal throttle time + is 1000ms. The throttle is only triggered under global memory pressure. + + With this interface, we can assign positive values for BATCH memcgs + and negative values for LS memcgs. Note that root memcg doesn't have + this file. + + memory.wmark_min_adj default value is 0, and inherit from its parent, + Note that the final effective wmark_min_adj will consider all the + hierarchical values, its value is the maximal(most conservative) + wmark_min_adj along the hierarchy but excluding intermediate default + values(zero). + + For example:: + + The following hierarchy + root + / \ + A D + / \ + B C + / \ + E F + + wmark_min_adj: A -10, B -25, C 0, D 50, E -25, F 50 + wmark_min_eadj: A -10, B -10, C 0, D 50, E -10, F 50 + + "echo xxx > memory.wmark_min_adj" set "wmark_min_adj". + "cat memory.wmark_min_adj" shows the value of "wmark_min_eadj". + +memory.exstat +============= + This file is under memory cgroup v1. + + memory.exstat stands for "extra/extended memory.stat", which is supposed + to provide hierarchical statistics. + + "wmark_min_throttled_ms" field is the total throttled time in milliseconds + due to positive memory.wmark_min_adj under global memory pressure. + +zombie memcgs reaper +==================== + After memcg was deleted, page caches still reference to this memcg + causing large number of dead(zombie) memcgs in the system. Then it + slows down access to "/sys/fs/cgroup/cpu/memory.stat", etc due to + tons of iterations, further causing various latencies. "zombie memcgs + reaper" is a tool to reclaim these dead memcgs. It has two modes: + + "Background kthread reaper" mode + -------------------------------- + In this mode, a kthread reaper keeps reclaiming at background, + some knobs are provided to control the reaper scan behaviour: + + - /sys/kernel/mm/memcg_reaper/scan_interval + + the scan period in second. Default is 5s. + + - /sys/kernel/mm/memcg_reaper/pages_scan + + the scan rate of pages per scan. Default 1310720(5GiB for 4KiB page). + + - /sys/kernel/mm/memcg_reaper/verbose + + output some zombie memcg information for debug purpose. Default off. + + - /sys/kernel/mm/memcg_reaper/reap_background + + on/off switch. Default "0" means off. Write "1" to switch it on. + + "One-shot trigger" mode + ----------------------- + In this mode, there is no guarantee to finish the reclaim, you may need + to check and launch multiple rounds as needed. + + - /sys/kernel/mm/memcg_reaper/reap + + users write "1" to trigger one round of zombie memcg reaping. -- 1.8.3.1