Squid Proxy "503 Service Unavailable" During Load Testing

[Anuska]

Hi ALL,

It's my first post here !

I use Google cloud and in my case egress traffic is allowed via squid proxy . Here my API details that's makes egress calls

API Usage Details: All requests are POST requests with JSON payloads.Responses are also Json each around 400 KB in size.Traffic Volume: The API handles around 300,000 requests per day.Concurrency: During load testing, the peak load is 15-20 concurrent users. I am using cloud run for my API.

When I perform load testing with 12 peak concurrent users, everything runs smoothly for about 5-6 minutes. However, after this period, Response time is slow and Squid begins to return " I/O error Unable to tunnel through proxy. Proxy returns "HTTP/1.1 503 Service Unavailable". This happens intermittently but frequently after 5-6 minutes and response time slows down as well

I have tried to disable caching initially since my data is dynamic so caching is no use.

I am new to GCP and squid . Any help would be appreciated.

Squid conf :

acl localnet src xxx.xx.xx.xx/xx

acl Safe_ports port 80        # http
acl Safe_ports port 443       # https
http_access deny !Safe_ports
http_access allow localhost manager
http_access deny manager

acl allowsites dstdomain "/etc/squid/approvedsites.lst"

acl exempt_ips src "/etc/squid/approved_ips.lst"

http_access allow allowsites           exempt_ips
#http_access allow all
http_access deny all

http_port 3128

coredump_dir /var/spool/squid
# refresh_pattern ^ftp:           1440   20%   10080
#refresh_pattern ^gopher:              #1440   0%    1440
#refresh_pattern -i (/cgi-bin/|\?) 0    0%    0
#refresh_pattern .               0      20%   4320
I tried only disable caching and I am reading and learning what more to do here. I increased the vm for squid to 4 cpu and 8 Gb. How many I for my usecase better the squid settings