Lots of networking errors from `from workos/base_client.rb:140:in 'block in WorkOS::BaseClient#execute_request` after upgrading the gem

[mintuhouse]

lib/openssl/ssl.rb:146:in 'OpenSSL::SSL::SSLContext#options=': can't modify frozen OpenSSL::SSL::SSLContext: #<OpenSSL::SSL::SSLContext:0x000077683dd27a18 @verify_mode=1, @verify_hostname=true, @cert_store=#<OpenSSL::X509::Store:0x000077685c61a468 @verify_callback=nil, @error=nil, @error_string=nil, @chain=nil>, @session_new_cb=#<Proc:0x000077683dd277c0 /railsapi/vendor/bundle/ruby/3.4.0/gems/net-http-0.9.1/lib/net/http.rb:1728>> (FrozenError)

NoMethodError: undefined method 'closed?' for nil lib/net/http.rb:2491:in 'Net::HTTP#begin_transport': undefined method 'closed?' for nil (NoMethodError)

openssl/buffering.rb:217:in 'OpenSSL::SSL::SSLSocket#sysread_nonblock': Bad file descriptor (Errno::EBADF)

lib/net/protocol.rb:229:in 'IO#wait_readable': stream closed in another thread (IOError)

net/http/response.rb:161:in 'Net::HTTPResponse.read_status_line': wrong status line: "0" (Net::HTTPBadResponse)

lib/net/protocol.rb:229:in 'IO#wait_readable': closed stream (IOError)

net/http/response.rb:628:in 'Net::HTTPResponse#read_chunked': wrong chunk size line: (Net::HTTPBadResponse)

net/http/response.rb:644:in 'Net::HTTPResponse#stream_check': attempt to read body out of block

The rate of these errors is quite high - though they are intermittent and succeed on retry
Guessing some sort of keep alive / connection pooling behaviour has changed in v9 (we were on v5 earlier)

If it helps this is from multi-threaded environment (solid queue)
And the base_url is https://auth.rhythms.ai

[gjtorikian]

Can you share any other info on APIs etc you are calling? (Like, I wonder what the source of this is, or whether it's all endpoints.)

[mintuhouse]

Its all over e.g.,
from workos/organization_membership_service.rb:159:in 'WorkOS::OrganizationMembershipService#update_organization_membership'

from workos/user_management.rb:756:in 'WorkOS::UserManagement#update_user'