Re: Socket exhaustion - a possible lead?
On 22 May 2009 as I do recall, Simon Smith wrote: > I'm still managing to trigger socket exhaustion from time to time - in fact > I think I'm getting better at it. The current sequence is reasonably > reproducible, provided you can find a site that will abort downloads on you > in mid-stream: > > 1 Load SktCounter > > 2 Start downloading some large files. Mp3s or something; enough to tax your > connection. I used the ISIHAC collection on archive.org. > > 3 Find a few image-heavy pages and load one. I suspect the faster your > connection the more pages (and the more image-heavy) you'll need. > > 4 Before the image-heavy page has finished loading, close it. > > 5 Load another image-heavy page, or the same one again. > > 6 At this stage, I still generally see SktCounter hovering at around 60-70 > connections. In fact, it either hovers at around 60-80, or it drops to zero; > with my setup I have never ever seen a positive socket count of less than > say, 60. Same here. > > 7 Keep loading pages and then interrupting the page downloads until one of > the large file downloads disconnects. This does sound very much like my frequent pattern of usage - and not infrequent experience of socket exhaustion. > > 8 As soon as the large downloads disconnect (NOT completing, but with bytes > to download remaining), I'm finding that my free socket count immediately > drops to 0. So no more internet until I've reset things. > > > I suspect a combination of an aborted page load (item 4) and one or more > disconnects (8) is the core of the problem. The only significance of the > image-heavy pages, I reckon, is that because they take longer to load it's > easier to interrupt them reliably at a critical point. This would also explain why the phenomenon was more frequent with a slower computer I'll see if I can duplicate your tests, but it might be tricky. Current settings: Max fetches: 24 Fetches per host: 5 Cached connections: 6 -- Harriet Bazley == Loyaulte me lie == What's the point in being grown up if you can't be childish sometimes?
Re: Quit and Restart Required
On 21 May 2009 as I do recall, Richard Porter wrote: > On 19 May 2009 list wrote: > > > On 20 May, Harriet Bazley wrote: > >> This sounds like the old 'singletasking while fetching stylesheets' > >> problem [...snip] In other words, you don't actually have to quit > >> NetSurf: just sit out the three minutes > > > Yes that's the syndrome. I can't close the offending window to stop the > > fetch, and thought I was quite silly when waiting that 60+ min. Now I see > > I have yet to experience silly ;) > > Wouldn't it be possible to abort the fetch by hitting Escape? > Not without specific Escape-detecting code inserted in that part of NetSurf, presumably. And as I assume the application thinks it's still multitasking in the desktop background at this point, that might be difficult -- Harriet Bazley == Loyaulte me lie == Do not underestimate the power of the Force.
Re: Socket exhaustion - a possible lead?
Simon Smith wrote: I'm now on the fresh 2.1 release, and the problem is as bad as ever. In particular, using a recent URL from the pull-down list, then selecting another recent URL /in the same window/ before the first page has loaded also seems to trigger it. So it's very easy by various means to accidentally get NetSurf in a state where the next aborted download will drop the socket count to zero. A nuisance. This doesn't affect all NetSurf users, and it doesn't affect all Iyonix users, therefore there must be something you are running which is affecting the network stack. Get a clean RISC OS 5 boot sequence and a clean NetSurf install, don't run anything else, and see if it still occurs. Cheers ---David -- Email: dr...@druck.org.uk Phone: +44-(0)7974 108301
Re: Socket exhaustion - a possible lead?
In message <6b7a725f50.zen44...@zen.co.uk> Simon Smith wrote: > In message <505f6e90cet...@netsurf-browser.org> > Michael Drake wrote: > > > In article , > >Simon Smith wrote: > > > > > > > I'm still managing to trigger socket exhaustion from time to time - in > > > fact I think I'm getting better at it. > > > > 1. What version of NetSurf are you using? This was a known problem with > >development builds for a period a couple of years back. > > > > 2. NetSurf allows the number of simultaneous connections it makes to be > >controlled. What values do you have in the Fetching section of > >Choices... Connection? > > > > 3. Try reducing the values. > > > > Michael > > I'm currently on build 7512 of 14/05/2009 > > I usually update every 1-3 weeks. > > Max fetches 24 > Fetches per host 20 > Cached connections 6 I'm now on the fresh 2.1 release, and the problem is as bad as ever. In particular, using a recent URL from the pull-down list, then selecting another recent URL /in the same window/ before the first page has loaded also seems to trigger it. So it's very easy by various means to accidentally get NetSurf in a state where the next aborted download will drop the socket count to zero. A nuisance. -- Simon Smith Where there's muck there's hope.
Re: Netsurf Export as Drawfile problem
In article , Dr Peter Young wrote: > On 23 May 2009 Dr Peter Young wrote: [Snip] > And all working now, in r7539. Many thanks again. > A slight oddity: the Save as Draw menu item gives one an HTML icon, > which changes to a Draw one when it's saved. Not a problem, but odd! > With best wishes, > Peter. I can confirm that Dev r7541 is Exporting Drawfiles okay. Icon in Menu now the correct Drawfile icon. Thanks Dave --
Re: Netsurf Export as Drawfile problem
On 23 May 2009 Dr Peter Young wrote: > On 23 May 2009 Dave Symes wrote: >> In article <1243088471.32028.15.ca...@duiker>, >>John-Mark Bell wrote: >>> The NetSurf developers are happy to announce the immediate availability >>> of NetSurf 2.1. >> Worth mentioning as a positive note. >> This 2.1 release Exports as a Drawfile okay. >> So obviously doesn't have the ongoing problem that the recent development >> versions have. > Yes, and if you look at the bug tracker, the bug is reported as fixed; > many thanks. However, it works with 2.1, as Dave says, but not yet > with the latest (at the time of writing) development build. I imagine > it will work with that soon. And all working now, in r7539. Many thanks again. A slight oddity: the Save as Draw menu item gives one an HTML icon, which changes to a Draw one when it's saved. Not a problem, but odd! With best wishes, Peter. -- Peter, \ / zfc Tm \ Prestbury, Cheltenham, Glos. GL52 Anne\/ ____\ England. and / / \ | | |\ | / _\ http://pnyoung.orpheusweb.co.uk family / \__/ \_/ | \| \__/ \__ pnyo...@ormail.co.uk