Skip to content

[Mellanox] Fix lua/orchagent storm detection race #697

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
lguohan merged 1 commit into from
Nov 30, 2018
+6 −3
Merged

[Mellanox] Fix lua/orchagent storm detection race #697

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
9 changes: 6 additions & 3 deletions orchagent/pfc_detect_mellanox.lua
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -71,6 +71,8 @@ for i = n, 1, -1 do
-- DEBUG CODE END.
(occupancy_bytes == 0 and packets - packets_last == 0 and (pfc_duration - pfc_duration_last) > poll_time * 0.8) then
if time_left <= poll_time then
redis.call('HDEL', counters_table_name .. ':' .. port_id, pfc_rx_pkt_key .. '_last')
Copy link
Copy Markdown
Contributor

@wendani wendani Nov 21, 2018 •
edited
Loading

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Need to understand how the change impacts the restore logic because restore relies on SAI_PORT_STAT_PFC_x_RX_PKTS_last, which is now reset on each storm detection. Because SAI_PORT_STAT_PFC_x_RX_PKTS_last is reset at a storm detection, the immediate next polling interval following the storm detection, the restore script can see pfc_rx_packets_last == 0, and skips the restore logic. https://github.com/Azure/sonic-swss/blob/6007e7f68cc103784208f69f8e6a0f12d4f2a193/orchagent/pfc_restore.lua#L42

So if there exists a situation that the storm restore is detected at the next polling interval of the storm detection occurrence, the restore signal is now delayed to 2 * polling intervals. Does this situation happen in realistic life, or its occurrence is just theorical?

Copy link
Copy Markdown
Contributor Author

@andriymoroz-mlnx andriymoroz-mlnx Nov 27, 2018

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

restoration script resets this field but will do it in proper time after orchagent reacted to the storm

if not ... pfc_wd_status ~= 'operational' ...
    ...
    redis.call('HSET', counters_table_name .. ':' .. port_id, pfc_rx_pkt_key .. '_last', pfc_rx_packets)
end

Copy link
Copy Markdown
Contributor

@wendani wendani Apr 2, 2019

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

If _RX_PKTS_last and _RX_PAUSE_DURATION_last is hdeleted at a storm signal, the earliest restore signal will be two polling intervals away from the time a detect signal is published.

redis.call('HDEL', counters_table_name .. ':' .. port_id, pfc_duration_key .. '_last')
redis.call('PUBLISH', 'PFC_WD', '["' .. KEYS[i] .. '","storm"]')
is_deadlock = true
time_left = detection_time
Expand All @@ -88,9 +90,10 @@ for i = n, 1, -1 do
-- Save values for next run
redis.call('HSET', counters_table_name .. ':' .. KEYS[i], 'SAI_QUEUE_STAT_PACKETS_last', packets)
redis.call('HSET', counters_table_name .. ':' .. KEYS[i], 'PFC_WD_DETECTION_TIME_LEFT', time_left)
redis.call('HSET', counters_table_name .. ':' .. port_id, pfc_rx_pkt_key .. '_last', pfc_rx_packets)
redis.call('HDEL', counters_table_name .. ':' .. port_id, pfc_duration_key .. '_last')
redis.call('HSET', counters_table_name .. ':' .. port_id, pfc_duration_key .. '_last', pfc_duration)
if is_deadlock == false then
redis.call('HSET', counters_table_name .. ':' .. port_id, pfc_rx_pkt_key .. '_last', pfc_rx_packets)
redis.call('HSET', counters_table_name .. ':' .. port_id, pfc_duration_key .. '_last', pfc_duration)
end
end
end
end
Expand Down