Hi James/Alan Thanks James for the explanation. As you said, the modem and 3G connection seem to work fine with Ubuntu 7.10. So that, I think, crosses off the USB modem or the 3G line as the cause of the error. The 3G line is supposed to have a max bandwidth of 7.2mbps, so I do not really understand how browsing normal web pages (www.sun.com or www.yahoo.com) would cause it to choke from the phone line POV. For the record though, pages like that of Google are loaded but it chokes with those that are not as bandwidth-light. So as James says, the whole thing seems to choke somewhere, but we do not know where that somewhere is...
Marc Torres On Fri, Jun 13, 2008 at 4:51 PM, James Carlson <[EMAIL PROTECTED]> wrote: > Alan DuBoff writes: >> After line 115 in your capture file, is that were it drops carrier? Up >> until then it seems to be talking to google.com, but it seems to get some >> 302 returns, I think those are ok. >> >> Not much I could see in a snoop file, seems the system log would be more >> helpful, at least to possibly narrow things down. > > The traces are hard to interpret, but (just for the record) I checked > through the kstats for the PPP kernel modules and the pppd user-space > logs, and some of the basic networking commands, and it's pretty clear > that the link itself (in terms of PPP and networking) is in good > shape. I also had him try some unusual fixes (mtu tweaks and ACCM) > just to make sure that there was nothing obviously wrong at the serial > level, despite the clean kstats. > > What seems to happen is that under some sort of stress (perhaps it > really *is* just HTTP traffic, but that hasn't really been proven to > my satisfaction), the link chokes up, and the only culprits I have > left are: > > - the USB subsystem > - the phone itself or its configuration > - the GSM network (always a dicey thing) > > Given those, and my limited ability to troubleshoot the phone itself > and the telephone network (sigh), I pointed him here. I'm hoping that > the USB driver has some sort of statistics that will show whether it's > performing correctly and perhaps whether there are flow control > problems. > > Under other OSes (Windows and Linux?), of course, he reports that it > runs fine. > > -- > James Carlson, Solaris Networking <[EMAIL PROTECTED]> > Sun Microsystems / 35 Network Drive 71.232W Vox +1 781 442 2084 > MS UBUR02-212 / Burlington MA 01803-2757 42.496N Fax +1 781 442 1677 > _______________________________________________ driver-discuss mailing list [email protected] http://mail.opensolaris.org/mailman/listinfo/driver-discuss
