From: Max Gurtovoy <maxg@mellanox.com>
To: David Miller <davem@davemloft.net>, <sagi@lightbitslabs.com>
Cc: <linux-nvme@lists.infradead.org>, <linux-block@vger.kernel.org>,
<netdev@vger.kernel.org>, <hch@lst.de>, <keith.busch@intel.com>
Subject: Re: [PATCH 10/11] nvmet-tcp: add NVMe over TCP target driver
Date: Sun, 18 Nov 2018 00:48:57 +0200 [thread overview]
Message-ID: <59335e55-7e19-3201-2f8b-beb681aba810@mellanox.com> (raw)
In-Reply-To: <20181117.121510.1754018619653365187.davem@davemloft.net>
On 11/17/2018 10:15 PM, David Miller wrote:
> From: Sagi Grimberg <sagi@lightbitslabs.com>
> Date: Thu, 15 Nov 2018 09:16:22 -0800
>
>> +static unsigned nvmet_tcp_recv_budget = 8;
>> +module_param_named(recv_budget, nvmet_tcp_recv_budget, int, S_IRUGO | S_IWUSR);
>> +MODULE_PARM_DESC(recv_budget, "recvs budget");
>> +
>> +static unsigned nvmet_tcp_send_budget = 8;
>> +module_param_named(send_budget, nvmet_tcp_send_budget, int, S_IRUGO | S_IWUSR);
>> +MODULE_PARM_DESC(send_budget, "sends budget");
>> +
>> +static unsigned nvmet_tcp_io_work_budget = 64;
>> +module_param_named(io_work_budget, nvmet_tcp_io_work_budget, int, S_IRUGO | S_IWUSR);
>> +MODULE_PARM_DESC(io_work_budget, "io work budget");
> I strongly suggest moving away from module parameters for this stuff.
agree here.
also, Sagi, can you explain about the performance trade-offs seen during
your development for these values ?
are they HCA/NIC dependent ?
should send/recv ratio be 1:1 ?
should total/send/recv ratio be 8:1:1 ?
>
> Create a genetlink socket family and allow run time configuration of these knobs
> by the user.
>
> Thanks.
WARNING: multiple messages have this Message-ID (diff)
From: maxg@mellanox.com (Max Gurtovoy)
Subject: [PATCH 10/11] nvmet-tcp: add NVMe over TCP target driver
Date: Sun, 18 Nov 2018 00:48:57 +0200 [thread overview]
Message-ID: <59335e55-7e19-3201-2f8b-beb681aba810@mellanox.com> (raw)
In-Reply-To: <20181117.121510.1754018619653365187.davem@davemloft.net>
On 11/17/2018 10:15 PM, David Miller wrote:
> From: Sagi Grimberg <sagi at lightbitslabs.com>
> Date: Thu, 15 Nov 2018 09:16:22 -0800
>
>> +static unsigned nvmet_tcp_recv_budget = 8;
>> +module_param_named(recv_budget, nvmet_tcp_recv_budget, int, S_IRUGO | S_IWUSR);
>> +MODULE_PARM_DESC(recv_budget, "recvs budget");
>> +
>> +static unsigned nvmet_tcp_send_budget = 8;
>> +module_param_named(send_budget, nvmet_tcp_send_budget, int, S_IRUGO | S_IWUSR);
>> +MODULE_PARM_DESC(send_budget, "sends budget");
>> +
>> +static unsigned nvmet_tcp_io_work_budget = 64;
>> +module_param_named(io_work_budget, nvmet_tcp_io_work_budget, int, S_IRUGO | S_IWUSR);
>> +MODULE_PARM_DESC(io_work_budget, "io work budget");
> I strongly suggest moving away from module parameters for this stuff.
agree here.
also, Sagi, can you explain about the performance trade-offs seen during
your development for these values ?
are they HCA/NIC dependent ?
should send/recv ratio be 1:1 ?
should total/send/recv ratio be 8:1:1 ?
>
> Create a genetlink socket family and allow run time configuration of these knobs
> by the user.
>
> Thanks.
next prev parent reply other threads:[~2018-11-17 22:50 UTC|newest]
Thread overview: 56+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-11-15 17:16 [PATCH 00/11] TCP transport binding for NVMe over Fabrics Sagi Grimberg
2018-11-15 17:16 ` Sagi Grimberg
2018-11-15 17:16 ` [PATCH 01/11] ath6kl: add ath6kl_ prefix to crypto_type Sagi Grimberg
2018-11-15 17:16 ` Sagi Grimberg
2018-11-15 17:16 ` [PATCH 02/11] iov_iter: introduce hash_and_copy iter helpers Sagi Grimberg
2018-11-15 17:16 ` Sagi Grimberg
2018-11-15 17:16 ` [PATCH 03/11] datagram: introduce skb_copy_and_hash_datagram_iter helper Sagi Grimberg
2018-11-15 17:16 ` Sagi Grimberg
2018-11-15 17:16 ` [PATCH 04/11] nvme-core: add work elements to struct nvme_ctrl Sagi Grimberg
2018-11-15 17:16 ` Sagi Grimberg
2018-11-17 22:18 ` Max Gurtovoy
2018-11-17 22:18 ` Max Gurtovoy
2018-11-15 17:16 ` [PATCH 05/11] nvmet: Add install_queue callout Sagi Grimberg
2018-11-15 17:16 ` Sagi Grimberg
2018-11-17 22:36 ` Max Gurtovoy
2018-11-17 22:36 ` Max Gurtovoy
2018-11-19 21:21 ` Sagi Grimberg
2018-11-19 21:21 ` Sagi Grimberg
2018-11-15 17:16 ` [PATCH 06/11] nvmet: allow configfs tcp trtype configuration Sagi Grimberg
2018-11-15 17:16 ` Sagi Grimberg
2018-11-17 22:38 ` Max Gurtovoy
2018-11-17 22:38 ` Max Gurtovoy
2018-11-15 17:16 ` [PATCH 07/11] nvme-fabrics: allow user passing header digest Sagi Grimberg
2018-11-15 17:16 ` Sagi Grimberg
2018-11-15 17:16 ` [PATCH 08/11] nvme-fabrics: allow user passing data digest Sagi Grimberg
2018-11-15 17:16 ` Sagi Grimberg
2018-11-15 17:16 ` [PATCH 09/11] nvme-tcp: Add protocol header Sagi Grimberg
2018-11-15 17:16 ` Sagi Grimberg
2018-11-15 17:16 ` [PATCH 10/11] nvmet-tcp: add NVMe over TCP target driver Sagi Grimberg
2018-11-15 17:16 ` Sagi Grimberg
2018-11-17 20:15 ` David Miller
2018-11-17 20:15 ` David Miller
2018-11-17 22:48 ` Max Gurtovoy [this message]
2018-11-17 22:48 ` Max Gurtovoy
2018-11-19 21:37 ` Sagi Grimberg
2018-11-19 21:37 ` Sagi Grimberg
2018-11-19 21:26 ` Sagi Grimberg
2018-11-19 21:26 ` Sagi Grimberg
2018-11-19 22:53 ` David Miller
2018-11-19 22:53 ` David Miller
2018-11-19 23:14 ` Sagi Grimberg
2018-11-19 23:14 ` Sagi Grimberg
2018-11-19 23:18 ` David Miller
2018-11-19 23:18 ` David Miller
2018-11-19 23:24 ` Sagi Grimberg
2018-11-19 23:24 ` Sagi Grimberg
2018-11-20 1:44 ` David Miller
2018-11-20 1:44 ` David Miller
2018-11-15 17:16 ` [PATCH 11/11] nvme-tcp: add NVMe over TCP host driver Sagi Grimberg
2018-11-15 17:16 ` Sagi Grimberg
2018-11-15 17:16 ` [PATCH nvme-cli 12/11] nvme: Add TCP transport Sagi Grimberg
2018-11-15 17:16 ` Sagi Grimberg
2018-11-15 17:16 ` [PATCH nvme-cli 13/11] fabrics: add tcp port tsas decoding Sagi Grimberg
2018-11-15 17:16 ` Sagi Grimberg
2018-11-15 17:16 ` [PATCH nvme-cli 14/11] fabrics: add transport header and data digest Sagi Grimberg
2018-11-15 17:16 ` Sagi Grimberg
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=59335e55-7e19-3201-2f8b-beb681aba810@mellanox.com \
--to=maxg@mellanox.com \
--cc=davem@davemloft.net \
--cc=hch@lst.de \
--cc=keith.busch@intel.com \
--cc=linux-block@vger.kernel.org \
--cc=linux-nvme@lists.infradead.org \
--cc=netdev@vger.kernel.org \
--cc=sagi@lightbitslabs.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.