On Mon, 8 Feb 2021, BERTRAND Joël wrote:
Can you turn off jumbo frames on wm0 and on the NAS? Seeing a window
size smaller than the frame size is a bit weird.
Today, I cannot. But when this nas was connected to a Realtek adapter,
I have the same throughput with a standard MTU (driver
On Mon, 8 Feb 2021, BERTRAND Joël wrote:
I have captured tcpdump output on both adapters (I post only one
second) :
wm0 (to euclide) :
14:06:09.014969 IP legendre.systella.fr.65508 >
euclide.systella.fr.iscsi-target: Flags [.], seq 107229888:107238836,
ack 78529, win 6528, options [nop
On Mon, 8 Feb 2021, Michael van Elst wrote:
On Mon, Feb 08, 2021 at 11:40:43AM +0100, BERTRAND Joël wrote:
If I do that in write mode, I suppose I will kill filesystem on
/dev/rdk5, isn't it ?
Definitely.
You need a test volume for a write test.
A not quite off-topic data-point he
RVP a écrit :
> On Mon, 8 Feb 2021, BERTRAND Joël wrote:
>
>> I have captured tcpdump output on both adapters (I post only one
>> second) :
>>
>> wm0 (to euclide) :
>> 14:06:09.014969 IP legendre.systella.fr.65508 >
>> euclide.systella.fr.iscsi-target: Flags [.], seq 107229888:107238836,
>> ac
On Mon, 8 Feb 2021, BERTRAND Joël wrote:
Does it store its cache on /opt?
No. Squid's cache is on /var (a local volume in Raid1).
OK.
Can you enable `noatime' and disable `log' when mounting /opt and
then run a dd read test on any 5GB+ file there (make sure the dd
cmd. has `ifla
Michael van Elst a écrit :
> On Mon, Feb 08, 2021 at 11:40:43AM +0100, BERTRAND Joël wrote:
>> Michael van Elst a écrit :
>>> For the "dd"-test of iSCSI, use the raw devices (not block devices,
>>> not files in a filesystem) and a block size of 1MB. Only if that
>>> works ok, it makes sense to look
On Mon, Feb 08, 2021 at 11:45:52AM +0100, BERTRAND Joël wrote:
> Some timestamps between lines seem strange :
> [ 313044.732208] ISCSI: scsipi_request GROW_RESOURCES
> [ 313044.832245] ISCSI: scsipi_request GROW_RESOURCES
> [ 313044.932283] ISCSI: scsipi_request GROW_RESOURCES
> [ 313045.032320]
Michael van Elst a écrit :
> On Mon, Feb 08, 2021 at 11:40:43AM +0100, BERTRAND Joël wrote:
>> Michael van Elst a écrit :
>>> For the "dd"-test of iSCSI, use the raw devices (not block devices,
>>> not files in a filesystem) and a block size of 1MB. Only if that
>>> works ok, it makes sense to look
On Mon, Feb 08, 2021 at 11:40:43AM +0100, BERTRAND Joël wrote:
> Michael van Elst a écrit :
> > For the "dd"-test of iSCSI, use the raw devices (not block devices,
> > not files in a filesystem) and a block size of 1MB. Only if that
> > works ok, it makes sense to look at filesystem I/O, i.e. your
Michael van Elst a écrit :
> The in-kernel initiator logs debug info when you set hw.iscsi.debug
> to a debug level > 0.
>
> 1 -> basic start/stop events, ioctls
> 5 -> more events, log commands and queue status
> 9 -> add handshake events
> 10 -> trace communication, task management, scsipi calls
Michael van Elst a écrit :
> For the "dd"-test of iSCSI, use the raw devices (not block devices,
> not files in a filesystem) and a block size of 1MB. Only if that
> works ok, it makes sense to look at filesystem I/O, i.e. your /opt
> filesystem.
If I do that in write mode, I suppose I wil
On Mon, Feb 08, 2021 at 10:09:50AM +, RVP wrote:
> Maybe, but, then there isn't any latency when running over loopback
> as all 3 of us (you, me & Michael) have seen. Plus, Michael mentioned
> getting a speed of 680 Mbits using iscsid over his network (maybe a
> 9.99.XX version of iscsid & ker
On Mon, Feb 08, 2021 at 10:13:16AM +0100, BERTRAND Joël wrote:
> I don't undestand how dd returns "no such file or directory" when I try
> :
>
> legendre# dd if=/dev/zero of=/opt/test iflag=direct oflag=direct
> count=5000 bs=1m
> dd: /opt/test: No such file or directory
oflag is the list
RVP a écrit :
> On Sun, 7 Feb 2021, BERTRAND Joël wrote:
>
>>> c) Does squid do anything there?
>>
>> Squid is used for proxy as my DSL connection is very poor.
>
> Does it store its cache on /opt?
No. Squid's cache is on /var (a local volume in Raid1).
>>> 2. Please run these co
Should be no problem. /emul/linux is just a link to /usr/pkg/emul/linux by
default, it can be separate partition of course or anything else, it
doesn't matter.
пн, 8 февр. 2021 г., 8:44 Thomas Mueller :
> Is it possible to run one's own little Linux in NetBSD by null-mounting
> directory or par
15 matches
Mail list logo
