Commit 7a895118 authored by Łukasz Nowak's avatar Łukasz Nowak

caddy-frontend: Workaround lost connections to haproxy

Due to unknown yet reason ATS and Caddy are loosing connections to haproxy,
which results with increased amount of 5xx (about 0.5% more cases).

No known solution was found yet on haproxy side, but this kind of workaround
stabilises frontend and definitely drops amount of in-frontend 5xx. The
observed situation with this workaround is no more 5xx sourcing in frontend
itself.
parent 9d810ee9
Pipeline #13002 failed with stage
in 0 seconds
...@@ -46,7 +46,7 @@ md5sum = 88af61e7abbf30dc99a1a2526161128d ...@@ -46,7 +46,7 @@ md5sum = 88af61e7abbf30dc99a1a2526161128d
[template-default-slave-virtualhost] [template-default-slave-virtualhost]
_update_hash_filename_ = templates/default-virtualhost.conf.in _update_hash_filename_ = templates/default-virtualhost.conf.in
md5sum = 266f175dbdfc588af7a86b0b1884fe73 md5sum = bd9e269130bac989faa639e0903814e2
[template-backend-haproxy-configuration] [template-backend-haproxy-configuration]
_update_hash_filename_ = templates/backend-haproxy.cfg.in _update_hash_filename_ = templates/backend-haproxy.cfg.in
...@@ -62,7 +62,7 @@ md5sum = 975177dedf677d24e14cede5d13187ce ...@@ -62,7 +62,7 @@ md5sum = 975177dedf677d24e14cede5d13187ce
[template-trafficserver-records-config] [template-trafficserver-records-config]
_update_hash_filename_ = templates/trafficserver/records.config.jinja2 _update_hash_filename_ = templates/trafficserver/records.config.jinja2
md5sum = 18076ae37306aca1f6ef11e2173c887f md5sum = a6a6e21036f8c56e69ad7ac8a1dfd656
[template-trafficserver-storage-config] [template-trafficserver-storage-config]
_update_hash_filename_ = templates/trafficserver/storage.config.jinja2 _update_hash_filename_ = templates/trafficserver/storage.config.jinja2
......
...@@ -43,6 +43,9 @@ ...@@ -43,6 +43,9 @@
timeout {{ slave_parameter['request-timeout'] }}s timeout {{ slave_parameter['request-timeout'] }}s
# force reset of X-Forwarded-For # force reset of X-Forwarded-For
header_upstream X-Forwarded-For {remote} header_upstream X-Forwarded-For {remote}
# workaround for lost connection to haproxy by reconnecting
try_duration 3s
try_interval 250ms
{%- endmacro %} {# proxy_header #} {%- endmacro %} {# proxy_header #}
{%- for tls in [True, False] %} {%- for tls in [True, False] %}
......
...@@ -80,10 +80,9 @@ CONFIG proxy.config.net.default_inactivity_timeout INT 86400 ...@@ -80,10 +80,9 @@ CONFIG proxy.config.net.default_inactivity_timeout INT 86400
# Origin server connect attempts. Docs: # Origin server connect attempts. Docs:
# https://docs.trafficserver.apache.org/records.config#origin-server-connect-attempts # https://docs.trafficserver.apache.org/records.config#origin-server-connect-attempts
############################################################################## ##############################################################################
# Try only once to connect (do not retry) # workaround for lost connection to haproxy by reconnecting
CONFIG proxy.config.http.connect_attempts_max_retries INT 0 CONFIG proxy.config.http.connect_attempts_max_retries INT 3
# Try only once with server marked dead (do not retry) CONFIG proxy.config.http.connect_attempts_max_retries_dead_server INT 1
CONFIG proxy.config.http.connect_attempts_max_retries_dead_server INT 0
CONFIG proxy.config.http.connect_attempts_rr_retries INT 3 CONFIG proxy.config.http.connect_attempts_rr_retries INT 3
CONFIG proxy.config.http.connect_attempts_timeout INT {{ ats_configuration['request-timeout'] }} CONFIG proxy.config.http.connect_attempts_timeout INT {{ ats_configuration['request-timeout'] }}
CONFIG proxy.config.http.post_connect_attempts_timeout INT {{ ats_configuration['request-timeout'] }} CONFIG proxy.config.http.post_connect_attempts_timeout INT {{ ats_configuration['request-timeout'] }}
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment