I'm using squid with a custom icap service. (Which code I plan to free)
--
http_port 3129 tproxy disable-pmtu-discovery=always
collapsed_forwarding on
dns_v4_first on
max_filedescriptors 8192
connect_retries 10
retry_on_error on
client_request_buffer_max_size 10250 KB
request_header_max_size 10240 KB
http_access allow manager localhost
http_access deny manager
http_access allow all
maximum_object_size 800 MB
maximum_object_size_in_memory 32 KB
cache_swap_low 90
cache_swap_high 95
#Server has 16Gb RAM
cache_mem 738 MB
cache_dir aufs /cache/sdb 228138269 16 256 min-size=1 max-size=838860800
buffered_logs off
icap_enable on
icap_send_client_ip on
icap_persistent_connections on
icap_preview_enable off
icap_206_enable off
icap_service service_reqmod reqmod_precache icap://127.0.0.1:50020/request bypass=0 max-conn=100 ipv6=off
icap_service service_respmod respmod_precache icap://127.0.0.1:50020/response bypass=0 max-conn=100 ipv6=off
acl html Content-Type -i html
adaptation_access service_respmod allow html
respmod_rep_header
adaptation_access service_reqmod allow all
# DEFAULT REFRESH PATTERNS
refresh_pattern -i (/cgi-bin/|?) 0 0% 0
refresh_pattern . 0 20% 4320
acl queries url_regex -i http://.*\?.*
acl queries url_regex -i http://.*/cgi-bin/.*
cache deny queries
cache allow all
client_persistent_connections on
server_persistent_connections on
debug_options ALL,0
I see the (squid-1) process RAM use slowly increasing. It's never going down. Ultil squid cannot allocate more memory and crashes.
Could be the icap client functionality ?
Alfrenovsky
_______________________________________________ squid-users mailing list squid-users@xxxxxxxxxxxxxxxxxxxxx http://lists.squid-cache.org/listinfo/squid-users
