[SPDK] Re: Unable to run OCF bdev, bdev not attached

SPDK bdev is in the user space. Looks like here you would like to have a try for the kernel version of OCF.

-----Original Message-----
From: Bob Chen <a175818323(a)gmail.com> 
Sent: Friday, August 28, 2020 16:01
To: Storage Performance Development Kit <spdk(a)lists.01.org>
Subject: [SPDK] Re: Unable to run OCF bdev, bdev not attached

I noticed SPDK has support of Linux AIO bdev, allowing us to read/write OS filesystem. Why couldn't we combine it with OCF?

Bob Chen <a175818323(a)gmail.com> 于2020年8月28日周五 下午3:54写道：

> Yes I've read this article. I'm just curious that is there any 
> possibility I can directly use the filesystem(ext4) above OS, while 
> entire filesystem (on HDD) is accelerated by SPDK + OCF transparently.
>
> I know SPDK has its BlobFS, however by far it only supports 
> append-only write. It seems to me that using the OS filesystem is a more practical way.
>
> Cao, Gang <gang.cao(a)intel.com> 于2020年8月28日周五 下午3:43写道：
>
>> >
>> > Are you using the OCF bdev for the FIO?
>> >
>>
>> No I'm not, the FIO command was directly run on /dev/sdk1. I suppose 
>> it's a aio bdev so I can read/write it directly with the OS?
>>
>> [Gang]: Looks like this is the problem. You will need to configure 
>> the SPDK OCF bdev for the FIO and the FIO is based on SPDK bdev 
>> layer. Please follow below link:
>>
>> https://mp.weixin.qq.com/s?__biz=MzI3NDA4ODY4MA==&mid=2653335424&idx=
>> 1&sn=0f4dfc6c884cea4780c016902dab59e8&chksm=f0cb5807c7bcd111a815ded1b
>> ab1dd8260167c59eab71ba729e27ec1f76b266e23b5f1a08bf2&token=1477511457&
>> lang=zh_CN#rd
>>
>> For the SPDK code, you still could use you current code, this is more 
>> about reference of the OCF and FIO configuration.
>>
>> -----Original Message-----
>> From: Bob Chen <a175818323(a)gmail.com>
>> Sent: Friday, August 28, 2020 15:38
>> To: Storage Performance Development Kit <spdk(a)lists.01.org>
>> Subject: [SPDK] Re: Unable to run OCF bdev, bdev not attached
>>
>> >
>> >
>> > What kind of workloads from FIO?
>>
>>
>> 4KB sequential write or randwrite
>>
>>
>> > And what's the capacity of the NVMe SSD as Cache device for the 
>> > Core device (HDD).
>> >
>>
>> 7TB SSD, 3.7TB HDD, 64KB OCF cache line
>>
>>
>> >
>> > Are you using the OCF bdev for the FIO?
>> >
>>
>> No I'm not, the FIO command was directly run on /dev/sdk1. I suppose 
>> it's a aio bdev so I can read/write it directly with the OS?
>>
>>
>> >
>> > If there is detailed configuration of FIO and OCF Bdev, that will 
>> > be helpful.
>> >
>> > Thanks,
>> > Gang
>> >
>> > -----Original Message-----
>> > From: Bob Chen <a175818323(a)gmail.com>
>> > Sent: Friday, August 28, 2020 15:20
>> > To: Storage Performance Development Kit <spdk(a)lists.01.org>
>> > Subject: [SPDK] Re: Unable to run OCF bdev, bdev not attached
>> >
>> > Hi, folks
>> >
>> > I solved this issue by myself. It's because the latest SPDK needs 
>> > 2MB hugepages while the spdk_tgt app applies 1GB by default.
>> >
>> > Then I began to run some fio commands on the core device 
>> > (/dev/sdk1), but surprisingly found that there was NO performance improvement at all.
>> >
>> > Using SPDK + OCF to accelerate a HDD by an NVMe SSD, is this a 
>> > correct scenario?
>> >
>> > <a175818323(a)gmail.com> 于2020年8月27日周四 下午5:58写道：
>> >
>> > > ./rpc.py bdev_nvme_attach_controller -b Nvme0 -t PCIe -a
>> > > 0000:3b:00.0 ./rpc.py bdev_aio_create /dev/sdk1 Aio0 512 ./rpc.py 
>> > > bdev_ocf_create
>> > > Cache0 wb Nvme0n1 Aio0 --cache-line-size 64
>> > >
>> > >
>> > > Inserting cache Cache0
>> > > Cache0: Metadata initialized
>> > > Cache0: Successfully added
>> > > Cache0: Cache mode : wb
>> > > Cache0: Attaching cache device failed
>> > > vbdev_ocf.c:1016:start_cache_cmpl: *ERROR*: Error -1000013 during 
>> > > start cache Cache0, starting rollback
>> > > Cache0: Cannot flush cache - cache device is detached
>> > > Cache0: Cache Cache0 successfully stopped
>> > >
>> > >
>> > > Looks like the cache/core device is not attached, what else 
>> > > should I do to make it work?
>> > >
>> > > P.S. I have run Open-CAS-Linux successfully on this machine, so 
>> > > the cas_cache and cas_cache ko are currently loaded. casctl and 
>> > > casadm tools exist as well.
>> > > _______________________________________________
>> > > SPDK mailing list -- spdk(a)lists.01.org To unsubscribe send an 
>> > > email to spdk-leave(a)lists.01.org
>> > >
>> > _______________________________________________
>> > SPDK mailing list -- spdk(a)lists.01.org To unsubscribe send an email 
>> > to spdk-leave(a)lists.01.org 
>> > _______________________________________________
>> > SPDK mailing list -- spdk(a)lists.01.org To unsubscribe send an email 
>> > to spdk-leave(a)lists.01.org
>> >
>> _______________________________________________
>> SPDK mailing list -- spdk(a)lists.01.org To unsubscribe send an email 
>> to spdk-leave(a)lists.01.org 
>> _______________________________________________
>> SPDK mailing list -- spdk(a)lists.01.org To unsubscribe send an email 
>> to spdk-leave(a)lists.01.org
>>
>
_______________________________________________
SPDK mailing list -- spdk(a)lists.01.org
To unsubscribe send an email to spdk-leave(a)lists.01.org