tomcat-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From kwirirai <kwiri...@designat7.co.zw>
Subject Re: Servlet thread safety in Tomcat
Date Tue, 10 Feb 2004 05:58:50 GMT
Thanks very much for your contribution guys! :-)
This has given me  new approaches to the problem.I intend to 
remodify,try the synchronization and do the testing/benchmarking.But 
this has made me think a lot about threading , it is really complicated 
, I have researched about it but never found a clean solution to 
multithreading in my apps.Initaly I thought with servlets I dont have to 
wory about the threading issue, now I know its everywhere.
I realy have todo a lot of testing.

Shapira, Yoav wrote:

>Howdy,
>
>  
>
>>Synchronization does produce overhead, but it's what you *must* do if
>>you will not re-write your servlets to no longer contain instance
>>fields, *and* you wish to provide thread safety.
>>    
>>
>
>Well-put.  And you might be pleasantly surprised at how small the
>overhead is.  Which is why I always suggest testing/benchmarking ;)
>
>  
>
>>There is a mode, and you have mentioned it before.. It's not a Tomcat
>>mode, it's that "Single threaded" thing. (see, I can't even write it
>>correctly, I don't even know what the exact name is).
>>    
>>
>
>It's SingleThreadModel, which is deprecated (with no replacement) in the
>latest version of the servlet specification.  So even if you're not
>using it already, don't start now.
>
>  
>
>>If the request that wants the 300 emails is taking "too long" (as
>>decided by the container), then that user's request thread is
>>    
>>
>suspended,
>  
>
>>while the other user's requests are handled.  That's the whole point of
>>multi-threading the servlet... (so that requests don't pile up)?
>>    
>>
>
>That's not done by tomcat: all requests have equal priority from the
>Tomcat point of view.  But the OS and JVM implement preemptive
>multitasking, and will split CPU time between the threads.  So you're
>right Mike, if request A takes a very long time, and request B starts
>after request A, request B does NOT have to wait until request A is done
>to get CPU time.
>
>Yoav Shapira
>
>
>
>This e-mail, including any attachments, is a confidential business communication, and
may contain information that is confidential, proprietary and/or privileged.  This e-mail
is intended only for the individual(s) to whom it is addressed, and may not be saved, copied,
printed, disclosed or used by anyone else.  If you are not the(an) intended recipient, please
immediately delete this e-mail from your computer system and notify the sender.  Thank you.
>
>
>---------------------------------------------------------------------
>To unsubscribe, e-mail: tomcat-user-unsubscribe@jakarta.apache.org
>For additional commands, e-mail: tomcat-user-help@jakarta.apache.org
>
>
>  
>


Mime
  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message