Re: [PATCH -next 4/5] blk-iocost: bypass if only one cgroup issues io

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

To: Yu Kuai <yukuai1@xxxxxxxxxxxxxxx>
Subject: Re: [PATCH -next 4/5] blk-iocost: bypass if only one cgroup issues io
From: Tejun Heo <tj@xxxxxxxxxx>
Date: Tue, 11 Oct 2022 22:04:05 -1000
Cc: axboe@xxxxxxxxx, linux-block@xxxxxxxxxxxxxxx, linux-kernel@xxxxxxxxxxxxxxx, yi.zhang@xxxxxxxxxx, "yukuai (C)" <yukuai3@xxxxxxxxxx>
In-reply-to: <c19008c3-512e-92cc-6be1-3ecc24e74341@huaweicloud.com>
References: <20221011083547.1831389-1-yukuai1@huaweicloud.com> <20221011083547.1831389-5-yukuai1@huaweicloud.com> <Y0WhpuqK/8CEZAGc@slm.duckdns.org> <c19008c3-512e-92cc-6be1-3ecc24e74341@huaweicloud.com>

On Wed, Oct 12, 2022 at 09:33:46AM +0800, Yu Kuai wrote:
> Perhaps a suitable configuration can avoid this problem.

Yeah, iocost is a throttling controller. The default parameters try to be
adaptive but it really needs benchmarked parameters to work properly.

Thanks.

-- 
tejun

References:
- [PATCH -next 0/5] blk-iocost: some random patches to improve iocost
  - From: Yu Kuai
- [PATCH -next 4/5] blk-iocost: bypass if only one cgroup issues io
  - From: Yu Kuai
- Re: [PATCH -next 4/5] blk-iocost: bypass if only one cgroup issues io
  - From: Tejun Heo
- Re: [PATCH -next 4/5] blk-iocost: bypass if only one cgroup issues io
  - From: Yu Kuai

Prev by Date: [RFC PATCHv2] mm: use stack_depot for recording kmemleak's backtrace
Next by Date: Re: [PATCH 1/1] virtio_pci: use common helper to configure SR-IOV
Previous by thread: Re: [PATCH -next 4/5] blk-iocost: bypass if only one cgroup issues io
Next by thread: [PATCH -next 2/5] blk-iocost: don't release 'ioc->lock' while updating params
Index(es):
- Date
- Thread

[Index of Archives] [Linux RAID] [Linux SCSI] [Linux ATA RAID] [IDE] [Linux Wireless] [Linux Kernel] [ATH6KL] [Linux Bluetooth] [Linux Netdev] [Kernel Newbies] [Security] [Git] [Netfilter] [Bugtraq] [Yosemite News] [MIPS Linux] [ARM Linux] [Linux Security] [Device Mapper]