Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > comp.programming.threads > #865 > unrolled thread
| Started by | "aminer" <aminer@videotron.ca> |
|---|---|
| First post | 2012-05-16 11:40 -0500 |
| Last post | 2012-05-16 11:55 -0500 |
| Articles | 4 — 1 participant |
Back to article view | Back to comp.programming.threads
About parallelhashlist.. "aminer" <aminer@videotron.ca> - 2012-05-16 11:40 -0500
Re: About parallelhashlist.. "aminer" <aminer@videotron.ca> - 2012-05-16 11:43 -0500
Re: About parallelhashlist.. "aminer" <aminer@videotron.ca> - 2012-05-16 11:47 -0500
Re: About parallelhashlist.. "aminer" <aminer@videotron.ca> - 2012-05-16 11:55 -0500
| From | "aminer" <aminer@videotron.ca> |
|---|---|
| Date | 2012-05-16 11:40 -0500 |
| Subject | About parallelhashlist.. |
| Message-ID | <jp0hqh$fog$1@dont-email.me> |
Hello, If you look at parallelhashlist in: http://pages.videotron.com/aminer/ Even if i use lock striping for the hash chains in parallelhashlist, synchronizing access to the counter reintroduces the scalability problems of exclusive locking. In this case, the counter is called a hot field because every mutative operation needs to access it. So my next step is to split the global counter to many local counters... also i am using a readewritelock in every chain, so my next step also is to lower the number of readwritelocks to consume also less memory... Thank you. Amine Moulay Ramdane.
[toc] | [next] | [standalone]
| From | "aminer" <aminer@videotron.ca> |
|---|---|
| Date | 2012-05-16 11:43 -0500 |
| Message-ID | <jp0hum$glv$1@dont-email.me> |
| In reply to | #865 |
I have wrote: > So my next step is to split the global counter to many local counters... I mean to split the global counter to many counters to enhance the scalability... Amine Moulay Ramdane. "aminer" <aminer@videotron.ca> wrote in message news:jp0hqh$fog$1@dont-email.me... > > > Hello, > > > > If you look at parallelhashlist in: > > http://pages.videotron.com/aminer/ > > > > Even if i use lock striping for the hash chains in parallelhashlist, > synchronizing > > access to the counter reintroduces the scalability problems of exclusive > locking. > > In this case, the counter is called a hot field because every mutative > operation needs > > to access it. So my next step is to split the global counter to many local > counters... > > also i am using a readewritelock in every chain, so my next step also is > to lower > > the number of readwritelocks to consume also less memory... > > > > Thank you. > > Amine Moulay Ramdane. > > > >
[toc] | [prev] | [next] | [standalone]
| From | "aminer" <aminer@videotron.ca> |
|---|---|
| Date | 2012-05-16 11:47 -0500 |
| Message-ID | <jp0i75$ib2$1@dont-email.me> |
| In reply to | #865 |
Hello, If you look at parallelhashlist in: http://pages.videotron.com/aminer/ Even if i use lock striping for the hash chains in parallelhashlist, synchronizing access to the counter reintroduces the scalability problems of exclusive locking. In this case, the counter is called a hot field because every mutative operation needs to access it. So my next step is to split the global counter to many counters to enhance the scalability... also i am using a readewritelock in every chain, so my next step also is to lower the number of readwritelocks to consume also less memory... Thank you. Amine Moulay Ramdane.
[toc] | [prev] | [next] | [standalone]
| From | "aminer" <aminer@videotron.ca> |
|---|---|
| Date | 2012-05-16 11:55 -0500 |
| Message-ID | <jp0imk$lom$1@dont-email.me> |
| In reply to | #867 |
I wrote: > Even if i use lock striping for the hash chains in parallelhashlist, > synchronizing access to the counter reintroduces the scalability > problems of exclusive locking. I was speaking about the counter that computes the number of the entries in the hashmap Amine Moulay Ramdane. "aminer" <aminer@videotron.ca> wrote in message news:jp0i75$ib2$1@dont-email.me... > > Hello, > > > If you look at parallelhashlist in: > > http://pages.videotron.com/aminer/ > > > Even if i use lock striping for the hash chains in parallelhashlist, > synchronizing access to the counter reintroduces the scalability > problems of exclusive locking. In this case, the counter is called > a hot field because every mutative operation needs to access it. > So my next step is to split the global counter to many counters to > enhance the scalability... also i am using a readewritelock in every > chain, > so my next step also is to lower the number of readwritelocks to consume > also less memory... > > > > Thank you. > > Amine Moulay Ramdane. > >
[toc] | [prev] | [standalone]
Back to top | Article view | comp.programming.threads
csiph-web