NETGEAR is aware of a growing number of phone and online scams. To learn how to stay safe click here.
Forum Discussion
GS716Tv2 with EAV features
Hi,
I'm having a terrible time debugging numerous problems with the switch and its AVB features provided by the EAV license.
First, and foremost, after running a level2 PTPv2 clock server on my main computer with a lower priority, both don't see each other and doesn't takeover as grandmaster. Thousands upon thousands of "bad header" packets are listed on the 802.1AS/Stats tab coming in on the port where I have the computer serving accurate clock. NIC is an Intel I210 with full hardware clock compliance as shown by:
$ ethtool -T eth2
Time stamping parameters for eth2:
Capabilities:
hardware-transmit (SOF_TIMESTAMPING_TX_HARDWARE)
hardware-receive (SOF_TIMESTAMPING_RX_HARDWARE)
hardware-raw-clock (SOF_TIMESTAMPING_RAW_HARDWARE)
PTP Hardware Clock: 0
Hardware Transmit Timestamp Modes:
off (HWTSTAMP_TX_OFF)
on (HWTSTAMP_TX_ON)
Hardware Receive Filter Modes:
none (HWTSTAMP_FILTER_NONE)
all (HWTSTAMP_FILTER_ALL)
PTP daemon is Linuxptp v1.3:
$ sudo ptp4l -mH2i eth2
ptp4l[166251.307]: selected /dev/ptp0 as PTP clock
ptp4l[166251.312]: driver changed our HWTSTAMP options
ptp4l[166251.312]: tx_type 1 not 1
ptp4l[166251.312]: rx_filter 1 not 12
ptp4l[166251.312]: port 1: INITIALIZING to LISTENING on INITIALIZE
ptp4l[166251.312]: port 0: INITIALIZING to LISTENING on INITIALIZE
ptp4l[166257.312]: port 1: LISTENING to MASTER on ANNOUNCE_RECEIPT_TIMEOUT_EXPIRES
ptp4l[166257.312]: selected best master clock a0369f.fffe.1cdd3b
ptp4l[166257.312]: assuming the grand master role
Second, Stream Reservation Protocol barely and rarely lets my AVB end-points connect to each other. Again, thousands of 'bad headers' are are shown on the MRP/MSRP_stats tab for the ports used by my end-points.
So the big questions is... does this latest x.13 firmware really do the full 802.1Q-2011 or not?
3 Replies
- I got the 'bad headers' problem fixed with 802.1AS by using P2P delay mode, but it refuses to even attempt to sync to the switch.
Apparently, the lines:Aug 11 12:50:39 bigmama ptp4l: [153971.512] driver changed our HWTSTAMP options
Aug 11 12:50:39 bigmama ptp4l: [153971.512] tx_type 1 not 1
Aug 11 12:50:39 bigmama ptp4l: [153971.512] rx_filter 1 not 12
means HWTSTAMP_FILTER_PTP_V2_L2_EVENT (linux/include/linux/net_tstamp.h) isn't available and falls back to HWTSTAMP_FILTER_ALL so it can't "hear" the switch.. but I can't confirm this guess just yet.
Still, though, the 802.1AS Port Settings page says the port the I210 is on isn't 802.1AS capable, which is a lie. - I went a few rounds with the Linuxptp developers and pretty much the conclusion is, this switch is hosed bad for standards compliance
- The helpful support agents got a beta firmware for me to try, and I have to say that it works. I've had an AVB stream running for 12 hours now off an alternate grandmaster. The 802.1AS Stats page is showing ports as capable with delay times, no bad headers. Very nice.
NETGEAR Academy
Boost your skills with the Netgear Academy - Get trained, certified and stay ahead with the latest Netgear technology!
Join Us!
