invalid arg to sysread deep within protocol.rb

[William E. Rubin]

I just made a script that fetches certain web pages, using Net::HTTP,
over and over in a loop. Everything worked fine, for about 5000
iterations, but then the following error was thrown (this is in Ruby
1.8.3):

C:/Ruby/lib/ruby/1.8/net/protocol.rb:133:in `sysread'
C:/Ruby/lib/ruby/1.8/net/protocol.rb:133:in `rbuf_fill'
C:/Ruby/lib/ruby/1.8/timeout.rb:45:in `timeout'
C:/Ruby/lib/ruby/1.8/timeout.rb:62:in `timeout'
C:/Ruby/lib/ruby/1.8/net/protocol.rb:132:in `rbuf_fill'
C:/Ruby/lib/ruby/1.8/net/protocol.rb:116:in `readuntil'
C:/Ruby/lib/ruby/1.8/net/protocol.rb:126:in `readline'
C:/Ruby/lib/ruby/1.8/net/http.rb:1988:in `read_status_line'
C:/Ruby/lib/ruby/1.8/net/http.rb:1977:in `read_new'
C:/Ruby/lib/ruby/1.8/net/http.rb:1046:in `request'
C:/Ruby/lib/ruby/1.8/net/http.rb:944:in `request_get'

The appropriate part in protocol.rb is this:

def rbuf_fill
timeout(@read_timeout) {
@rbuf << @io.sysread(1024)
}
end

So sysread suddenly considers "1024" to be an invalid number of bytes
to read? After having considered it valid on each of the previous 5000
iterations?

Any clues? Thanks.

 

[Logan Capaldo]

I just made a script that fetches certain web pages, using Net::HTTP,
over and over in a loop. Everything worked fine, for about 5000
iterations, but then the following error was thrown (this is in Ruby
1.8.3):

C:/Ruby/lib/ruby/1.8/net/protocol.rb:133:in `sysread'
C:/Ruby/lib/ruby/1.8/net/protocol.rb:133:in `rbuf_fill'
C:/Ruby/lib/ruby/1.8/timeout.rb:45:in `timeout'
C:/Ruby/lib/ruby/1.8/timeout.rb:62:in `timeout'
C:/Ruby/lib/ruby/1.8/net/protocol.rb:132:in `rbuf_fill'
C:/Ruby/lib/ruby/1.8/net/protocol.rb:116:in `readuntil'
C:/Ruby/lib/ruby/1.8/net/protocol.rb:126:in `readline'
C:/Ruby/lib/ruby/1.8/net/http.rb:1988:in `read_status_line'
C:/Ruby/lib/ruby/1.8/net/http.rb:1977:in `read_new'
C:/Ruby/lib/ruby/1.8/net/http.rb:1046:in `request'
C:/Ruby/lib/ruby/1.8/net/http.rb:944:in `request_get'

The appropriate part in protocol.rb is this:

def rbuf_fill
timeout(@read_timeout) {
@rbuf << @io.sysread(1024)
}
end

So sysread suddenly considers "1024" to be an invalid number of bytes
to read? After having considered it valid on each of the previous 5000
iterations?

Any clues? Thanks.

Maybe there aren't 1024 bytes left? According to the docs sysread is
supposed to raise EOFError at the end of the file, are you sure
that's not being raised?

 

[William E. Rubin]

Maybe there aren't 1024 bytes left? According to the docs sysread is

supposed to raise EOFError at the end of the file, are you sure
that's not being raised?

Yes, I'm sure. Now that I look back at my original post, I notice that
I left out a crucial piece of information:

The error raised was an "Invalid Argument" error.

 

[William E. Rubin]

This seems to occur every few thousand iterations (not the same number
of iterations every time).