You can push them pretty hard. We're reliably running at an average of around
90% wire speed on all four links for MeerKAT. But be sure to factor-in the
overheads of the core...
In our case:
L1 Ethernet HDR & FTR44B
IPv4 header 20B
UDP header8B
Application heade
Oh, yeah, the good old "obvious" problem! :-)
Make sure the MTU of all systems involved (including any intervening switches!)
is large enough to handle the output packet size (plus overhead). For local
networks with PCs and ROACH2s, we usually use an MTU of 9000 everywhere.
Watch out if you ar
*Solution*
It turns out that the basic problem was that I was sending 4096 bytes per
frame to the one_GbE block. After consulting with an expert on internet
protocols, it was suggested that I change the packet size from 4096 to 1024
bytes. Once I did this, I can clearly see the packets arriving
Marc,
Here is an update on the issue:
Sometimes when I issue a reset of the one_GbE yellow block, the V6 port
will get into a mode where it does send data out of the port. I have a few
telemetry points connected to the front panel, so I know that when it is in
this state, the following is true:
I have been getting a lot of helpful suggestions and will respond to
everything here:
1) Throughout all of today's testing, I have ensured that the tx_overrun
flag is never set.
2) On all runs that I have tried, the tx_afull flag is always set prior to
issuing the tap_start command.
3) I have the
Hi all,
I have an application that generates 40 Gb/s of data and I'd like to
output it over 4 x 10GbE links (i.e. one roach2 mezzanine card).
I'm going to have to throw some data away to make room for packetising
overhead, and right now I'm trying to work out how much, to see
whether I should jus
Hi Joe, Casperites,
You want to prevent the tx buffer overflowing at all costs when it
comes to the 10GbE block. Try issuing a reset by toggling the gbe reset port on
the yellow block. that usually fixes things for me and starts it right back up.
I don’t think I’ve ever needed to re-is
Hi Joe, all
I ran into some issues on a roach 1 where the 10 gbe buffer would fill
up and overflow before i had run tap_start (if I recall correctly). In
that case, no data would ever be transmitted, and I had to add a
software register to the overflow port to see this was the case.
Executing t
Hey Joe, Casperites,
Can you log into the powerpc and run ifconfig on a roach 2? I've done this
to troubleshoot 10 gbe on a roach1, and I was able to see that I truly did
enable the ten gbe using tap start, as it is listed in the ifconfig output
and then there is some data about packets sent and
Hello
> ?tap-info Riometer0
>
> #log info 964298200737 raw peer\_02:02:00:00:00:00\_at\_29
> #log info 964298200737 raw peer\_d0:67:e5:f1:9b:e7\_at\_15000
> #log info 964298200738 raw peer\_00:1a:a0:b1:1e:9b\_at\_15000
> #log info 964298200738 raw peer\_00:1e:4f:cc:cc:b4\_at\_15000
So that is a g
Jason,
They are aligned now, but still no dice. Is the expected behavior that the
one_GbE will broadcast a UDP packet once the EOF and last valid word flags
are set?
-Joe Kujawski
Are your EOFs aligned with the last valid word into the core (ie on the same
clock cycle) ? This is typically the issue that causes this behaviour.
On 13 Mar 2014, at 16:58, Kujawski, Joseph wrote:
> What event causes the one_GbE yellow block to send data to the tx_destip?
>
> I have a simp
What event causes the one_GbE yellow block to send data to the tx_destip?
I have a simple design that fills tx_data with a count until tx_afull goes
high. Once I enable the ethernet (with the ?tap-start command in KATCP), I
can ping the port, however, I have determined that the V6 port of the
Roa
Mark,
First I am now pinging the ROACH2 V6 port, but I still get the log info
lines on the KATCP screen. Executing the ?tap-info produces this result:
#log info 964298199431 raw
detected\_time\_warp\_or\_stall,\_rescheduling\_periodic\_task\_for\_964298199.434018s
#log info 964298200011 raw
dete
Hello
On Wed, Mar 12, 2014 at 7:31 PM, Kujawski, Joseph wrote:
> Jack,
>
> Thanks. For reference, the way to enter the command in the KATCP directly
> (at least in my version) is:
>
> ?tap-start Riometer0 one_GbE 192.168.6.29 4321 02:02:00:00:00:00
>
> Where
> Riometer0 - name of the device
> on
15 matches
Mail list logo