Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 10 Dec 2018 11:05:09 +0500
From:      "Eugene M. Zheganin" <eugene@zhegan.in>
To:        dtrace@freebsd.org
Subject:   Re: iotop for iSCSI or zfs datasets
Message-ID:  <5b23c43b-a00f-0703-60e1-00920d5dbe79@zhegan.in>
In-Reply-To: <CAOtMX2gEnzzq9XZQqU2WLLGxAYNhkM7C8GGqM03X=eemK9Wqhw@mail.gmail.com>
References:  <67bc8f0a-8702-c6ee-9d37-3a5064709126@zhegan.in> <CAOtMX2gEnzzq9XZQqU2WLLGxAYNhkM7C8GGqM03X=eemK9Wqhw@mail.gmail.com>

next in thread | previous in thread | raw e-mail | index | archive | help
Hello,

On 07.12.2018 20:25, Alan Somers wrote:
>
>> I also can see the iotop in the context of network data transferred
>> using trafshow, but this doesn't scale well to iops, and I would see
>> reads/writes per second in terms that block devices use.
>>
>>
>> Thanks.
>>
>> Eugene.
> What's wrong with gstat?  Use "gstat -a" to suppress idle devices.
> -Alan

Everything is wrong with gstat in this case.

First of all, this is what's wrong - 
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=227465 , and gstat 
required zvols to be volmode=geom to use it this way.

Second, volmode=dev is supposed to be more performing.

Finally, I find hilarious that first answer in the dtrace@ was "don't 
use dtrace".

Eugene.




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?5b23c43b-a00f-0703-60e1-00920d5dbe79>