Rajkumar-Marvell

> applyView: current view switches: 0 != temporary view switches: 1, FATAL > this suggest that there was not switch in previous run of syncd, > what scenario is this...

> after reaching this state, could you check "redis-cli -n 1" and "keys *" to see if there are any ASIC_STATE keys except TEMP_ASIC_STATE ? Attached redis-cli output in issue...

> in ASIC.txt there is no ASIC_STATE* entries, this seems like those entreis were removed or database was cleared, could we work together on that this week? Yes. We can...

> Since ASIC.txt dump don't contain "HIDDEN" and any ASIC_STATE* keys, then this means database was flushed, > also i suspect RIDTOVID and VIDTORID hashes to be also small count...

> could you try to execute "sudo systemctl restart swss" or "config reload" instead of "syncd" ?, i think currently restart syncd is not supported. > is there are reason...

@yxieca @kcudnik Hi, Is this bug scheduled for fix ? Let me know if you need any more logs to fix this issue. Thanks.

> Quick comment regarding choice of Option 1 and Option 2. Option 1: Doesn't provide granularity of enabling and disabling at granular level. This approach also means that all the...

> Instead od adding same attributes for all objects would it be better to add global API for this ? Hi @kcudnik, If we have to manage this by global...

> Instead of adding 2 attributes to all objects add global API that will query those values, no need for definig which objects support that, if some will not support...

@chuzirui, We need 8_LANE support for Marvell ASIC. Do you plan to discuss this PR in community meeting ?