Catalin, Salgau, hi. Did you send a patch for this packet-size-tuning feature? It seems like it would be a nice patch for contrib/ if you could put a nice description at the top, including its motivation and your personal experiences. During testing you might even get a chance to jot down some details about specific performance differences, to motivate potential users to try the patch. (I'm composing this in Thunderbird, and I hope I'm not going to send HTML mail!) On 06/14/2014 11:59 PM, Catalin Salgau wrote: > On 15/06/2014 4:06 AM, Ed Cashin wrote: >> Hi, Catalin Salgau. >> >> I have questions below between selected quotes. >> >> On 2014-06-10 07:59, Catalin Salgau wrote: >>> I would like to request two changes before release. >>> - An option to restrict the size of packets over automatic detection of >>> MTU. >> You mean like if the MTU is 9000, you want the ability to tell the >> vblade to act like it's smaller, right? > Yes. That's the gist of it. > I believe there is some value in the ability to manually tweak the > maximum packet size used by vlade. > At the very least it would help with determining optimal parameters for > a deployment/use case. >> If you have some numbers to share (MTUs and packet sizes as well as >> throughput rates and latencies), that would fill out your interesting >> story with important details. > I sadly made no effort to document it, and, in retrospect, it might have > made for an interesting study.. > As 'methodology', the target was configured to support 9014 byte frames > and the initiator was switched between the two tested packet sizes on > Windows XP x86 and Windows 7 amd64. > Due to some problems with making these changes to multiple test images, > I haven't replicated the results over a larger set of machines. > I'm hoping to get back to this in a week or two. >> ... >>> - change to vblade(8) manual Synopsis section to include current syntax >> That change might be simple enough for a patch to be easier for me to >> understand than a description, so please send the patch if you don't >> mind. > This was probably poorly worded. > What I was trying to say was that the Synopsis section in the manual > page supplied with vblade has not been kept in sync with options added > to vblade. > Since changing this as suggested below would yield a line longer than 80 > chars, I'm not providing a real patch; I don't know how one should > format this. Use it as reference, maybe? > > --- a/vblade.8 > +++ b/vblade.8 > @@ -6,1 +6,1 @@ > -.B vblade [ -m mac[,mac...] ] shelf slot netif filename > +.B vblade [-b bufcnt] [-o offset] [-l length] [-dsr] [ -m mac[,mac...] > ] shelf slot netif filename > > ------------------------------------------------------------------------------ > HPCC Systems Open Source Big Data Platform from LexisNexis Risk Solutions > Find What Matters Most in Your Big Data with HPCC Systems > Open Source. Fast. Scalable. Simple. Ideal for Dirty Data. > Leverages Graph Analysis for Fast Processing & Easy Data Exploration > http://p.sf.net/sfu/hpccsystems > _______________________________________________ > Aoetools-discuss mailing list > Aoetools-discuss@xxxxxxxxxxxxxxxxxxxxx > https://lists.sourceforge.net/lists/listinfo/aoetools-discuss -- Ed ------------------------------------------------------------------------------ _______________________________________________ Aoetools-discuss mailing list Aoetools-discuss@xxxxxxxxxxxxxxxxxxxxx https://lists.sourceforge.net/lists/listinfo/aoetools-discuss
