[opensuse] unable to access website after 10.2 install
This has me scratching my head. After a clean install of 10.2, I am now unable to access my work website (www.marymount.edu). All other websites load up just fine. At first, I thought it was a problem with this specific website, but if I dual boot to windows, the site comes up just fine, and if I plug my laptop (currently running suse 10.1) to the Ethernet, the site also loads up just fine. So, I imagine something within 10.2 is keeping it from loading--but not sure where the problem is. I've disabled firewall, ipv6, but still cannot load the site. If I ping the site, I get: chip@linux-19nx:~> ping -c 3 marymount.edu http://marymount.edu/ PING marymount.edu http://marymount.edu/ (198.100.0.33 http://198.100.0.33/) 56(84) bytes of data. --- marymount.edu http://marymount.edu/ ping statistics --- 3 packets transmitted, 0 received, 100% packet loss, time 2001ms and traceroute give me: linux-19nx:~ # traceroute marymount.edu http://marymount.edu/ traceroute to marymount.edu http://marymount.edu/ ( 198.100.0.33 http://198.100.0.33/), 30 hops max, 40 byte packets 1 192.168.15.1 http://192.168.15.1/ ( 192.168.15.1 http://192.168.15.1/) 0.744 ms 0.386 ms 0.396 ms 2 ip24-250-96-1.dc.dc.cox.net http://ip24-250-96-1.dc.dc.cox.net/ ( 24.250.96.1 http://24.250.96.1/) 18.913 ms 25.217 ms 23.428 m 3 mrfdaggc01.dc.dc.cox.net http://mrfdaggc01.dc.dc.cox.net/ ( 68.100.1.65 http://68.100.1.65/) 16.584 ms 12.408 ms 13.617 ms 4 ip68-100-0-1.dc.dc.cox.net http://ip68-100-0-1.dc.dc.cox.net/ (68.100.0.1 http://68.100.0.1/) 14.263 ms 9.064 ms 11.708 ms 5 mrfddsrj01gex070004.rd.dc.cox.net http://mrfddsrj01gex070004.rd.dc.cox.net/ (68.100.0.161 http://68.100.0.161/) 13.285 ms 13.319 ms 1 .714 ms 6 68.1.0.251 http://68.1.0.251/ ( 68.1.0.251 http://68.1.0.251/) 16.000 ms 14.546 ms 16.563 ms 7 68.105.31.74 http://68.105.31.74/ (68.105.31.74 http://68.105.31.74/) 16.595 ms 15.889 ms 15.372 ms 8 v3495.mpd03.jfk02.atlas.cogentco.com http://v3495.mpd03.jfk02.atlas.cogentco.com/ ( 154.54.6.45 http://154.54.6.45/) 19.512 ms 20.360 ms * 9 * * t9-4.mpd01.dca01.atlas.cogentco.com http://t9-4.mpd01.dca01.atlas.cogentco.com/ ( 154.54.6.138 http://154.54.6.138/) 21.644 ms 10 * * * 11 * * * 12 * * * 13 * * * 14 * * * 15 * * * 16 * * * 17 * * * 18 * * * 19 * * * 20 * * * 21 * * * 22 * * * 23 * * * 24 * * * 25 * * * 26 * * * 27 * * * 28 * * * 29 * * * 30 * * * Anyone have an idea what is going on with 10.2 that may be causing this? Every other site I try to load comes up just fine. Would appreciate any suggestions--I need to access this site for my job. Thanks! Chip -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Mandag 18 december 2006 04:59 skrev Chip Cooper:
te (www.marymount.edu). All othe
Good morning, - FWIW, I can't ping nor see the website from my place either... (Denmark) - using 10.2 -- ------------------------------ Med venlig hilsen/Best regards Verner Kjærsgaard -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Le dimanche 17 décembre 2006 19:59, Chip Cooper a écrit : It doesn't help you much, but I can comfirm your observations. Doesn't work in openSUSE 10.2, does work on a Windows PC. A Windows XP Guest on openSUSE 10.2 (vmware) can also connect to the website. And finally a connection from Fedora Core 6 fails. Maybe a kernel problem? -- Gruß & Happy Holidays Andreas -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On Mon, 2006-12-18 at 00:13 -0800, Andreas wrote:
Le dimanche 17 décembre 2006 19:59, Chip Cooper a écrit : It doesn't help you much, but I can comfirm your observations. Doesn't work in openSUSE 10.2, does work on a Windows PC.
A Windows XP Guest on openSUSE 10.2 (vmware) can also connect to the website.
And finally a connection from Fedora Core 6 fails.
Maybe a kernel problem?
Maybe a web browser issue? Which browsers are working and which are not working? -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Maybe a web browser issue? Which browsers are working and which are not working?
Well, I've tried konqueror, firefox, mozilla, and opera. None work. Can't get a ping from the terminal. Don't think it is a browser issue. It is interesting Andreas was unable to load the site using Fedora core 6. So maybe a kernel issue/parameter/bug? No idea what is going on. I'm open to ideas! -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On 2006/12/18 09:22 (GMT-0500) Chip Cooper apparently typed:
Well, I've tried konqueror, firefox, mozilla, and opera. None work. Can't get a ping from the terminal. Don't think it is a browser issue. It is interesting Andreas was unable to load the site using Fedora core 6. So maybe a kernel issue/parameter/bug? No idea what is going on. I'm open to ideas!
As many failures as seen in this thread, it has to be something about the site configuration, likely IPV6. What was that URL again? -- "Let your conversation be always full of grace." Colossians 4:6 NIV Team OS/2 ** Reg. Linux User #211409 Felix Miata *** http://mrmazda.no-ip.com/ -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On 2006/12/18 09:22 (GMT+-0500) Chip Cooper apparently typed: >>>http://www.marymount.edu/ > Well, I've tried konqueror, firefox, mozilla, and opera. None work. > Can't get a ping from the terminal. Don't think it is a browser issue. > It is interesting Andreas was unable to load the site using Fedora core > 6. So maybe a kernel issue/parameter/bug? No idea what is going on. I'm > open to ideas! On same box I have XP SP2, Factory/10.2 final, 9.3, 10.0, Fedora 6, Debian Etch and Kubuntu 6.10. Only XP, Kubuntu, 10.0 & 9.3 can reach it on that box. Other boxes with 10.0 & OS/2 have no problem either. MarymountUniversity.demarc.cogentco.com is the last point on traceroute reached by 10.2. -- "Let your conversation be always full of grace." Colossians 4:6 NIV Team OS/2 ** Reg. Linux User #211409 ** Rotary ONLY since 1973 Felix Miata *** http://mrmazda.no-ip.com/ -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Felix Miata wrote:
On 2006/12/18 09:22 (GMT+-0500) Chip Cooper apparently typed:
Well, I've tried konqueror, firefox, mozilla, and opera. None work. Can't get a ping from the terminal. Out of curiosity (and frustration), I decided to install the 2.6.16 kernel from my 10.1 disks. When I boot with this kernel, the website loads fine. So, it seems something is amiss in 2.6.18 in allowing certain sites to load, but I have no clue what it could be.
Chip -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On 2006/12/18 22:38 (GMT-0500) Chip Cooper apparently typed:
Well, I've tried konqueror, firefox, mozilla, and opera. None work. Can't get a ping from the terminal.
Out of curiosity (and frustration), I decided to install the 2.6.16 kernel from my 10.1 disks. When I boot with this kernel, the website loads fine. So, it seems something is amiss in 2.6.18 in allowing certain sites to load, but I have no clue what it could be.
Might be time to take this one to bugzilla. -- "Let your conversation be always full of grace." Colossians 4:6 NIV Team OS/2 ** Reg. Linux User #211409 Felix Miata *** http://mrmazda.no-ip.com/ -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On 12/19/06, Chip Cooper
Felix Miata wrote:
On 2006/12/18 09:22 (GMT+-0500) Chip Cooper apparently typed:
Well, I've tried konqueror, firefox, mozilla, and opera. None work. Can't get a ping from the terminal. Out of curiosity (and frustration), I decided to install the 2.6.16 kernel from my 10.1 disks. When I boot with this kernel, the website loads fine. So, it seems something is amiss in 2.6.18 in allowing certain sites to load, but I have no clue what it could be.
Chip
-- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Out of curiosity tried it from the office from ThinkPad with OpenSuSE 10.2 (kernel 2.6.18.2-34-default) and site opened in FF without any problem. In both cases (in the office and at home) I'm behind NAT/firewall. Since initial report says ping to this address did not work, it's not browser, but something below. -- Mark Goldstein -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On Monday 18 December 2006 22:33, Mark Goldstein wrote:
Since initial report says ping to this address did not work, it's not browser, but something below.
You can not attribute any significance to ping not working. Many sites block ping. -- _____________________________________ John Andersen
On Monday 18 December 2006 22:33, Mark Goldstein wrote:
Out of curiosity tried it from the office from ThinkPad with OpenSuSE 10.2 (kernel 2.6.18.2-34-default) and site opened in FF without any problem.
What kind of Nic does the Thinkpad have? My test was on a brand new dell server, with the same kernal as yours (but x86_64), using Broadcom Corporation NetXtreme BCM5721 Gigabit Ethernet PCI Express -- _____________________________________ John Andersen
On 12/19/06, John Andersen
On Monday 18 December 2006 22:33, Mark Goldstein wrote:
Out of curiosity tried it from the office from ThinkPad with OpenSuSE 10.2 (kernel 2.6.18.2-34-default) and site opened in FF without any problem.
What kind of Nic does the Thinkpad have?
eth0 device: Intel Corporation 82801CAM (ICH3) PRO/100 VE (LOM) Ethernet Controller -- Mark Goldstein -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
John Andersen wrote:
On Monday 18 December 2006 22:33, Mark Goldstein wrote:
Out of curiosity tried it from the office from ThinkPad with OpenSuSE 10.2 (kernel 2.6.18.2-34-default) and site opened in FF without any problem.
What kind of Nic does the Thinkpad have?
My test was on a brand new dell server, with the same kernal as yours (but x86_64), using Broadcom Corporation NetXtreme BCM5721 Gigabit Ethernet PCI Express
Is it an arch problem? My 10.2 here at home (x86_64) cannot reach this site either, BUT my 10.2 machine at work, which is i386, did work. I would assume Mark's Thinkpad is an i386 arch as well, and John you mentioned yours was x86_64, so the apparent pattern that I am beginning to see is not a 10.2 problem but a 10.2 x86_64 problem. -- Joe Morris Registered Linux user 231871 running openSUSE 10.2 x86_64 -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On 12/19/06, Joe Morris (NTM)
John Andersen wrote:
On Monday 18 December 2006 22:33, Mark Goldstein wrote:
Is it an arch problem? My 10.2 here at home (x86_64) cannot reach this site either, BUT my 10.2 machine at work, which is i386, did work. I would assume Mark's Thinkpad is an i386 arch as well, and John you mentioned yours was x86_64, so the apparent pattern that I am beginning to see is not a 10.2 problem but a 10.2 x86_64 problem.
Yes, Thinkpad is i386 as well as IBM Aptiva at home that also had no problem accessing this site... -- Mark Goldstein -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On Tue, 2006-12-19 at 18:26 +0800, Joe Morris (NTM) wrote:
John Andersen wrote:
On Monday 18 December 2006 22:33, Mark Goldstein wrote:
Out of curiosity tried it from the office from ThinkPad with OpenSuSE 10.2 (kernel 2.6.18.2-34-default) and site opened in FF without any problem.
What kind of Nic does the Thinkpad have?
My test was on a brand new dell server, with the same kernal as yours (but x86_64), using Broadcom Corporation NetXtreme BCM5721 Gigabit Ethernet PCI Express
Is it an arch problem? My 10.2 here at home (x86_64) cannot reach this site either, BUT my 10.2 machine at work, which is i386, did work. I would assume Mark's Thinkpad is an i386 arch as well, and John you mentioned yours was x86_64, so the apparent pattern that I am beginning to see is not a 10.2 problem but a 10.2 x86_64 problem.
My laptop is i386 and cannot reach the site, so no it is not an arch problem. -- Ken Schneider UNIX since 1989, linux since 1994, SuSE since 1998 -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Joe Morris (NTM) wrote:
John Andersen wrote:
...
Is it an arch problem? My 10.2 here at home (x86_64) cannot reach this site either, BUT my 10.2 machine at work, which is i386, did work. I would assume Mark's Thinkpad is an i386 arch as well, and John you mentioned yours was x86_64, so the apparent pattern that I am beginning to see is not a 10.2 problem but a 10.2 x86_64 problem.
suse 10.2 x86_64, Rhine II: can't reach www.marymount.edu suse 10.2 Intel Core Duo, Intel PRO/100 VE: can't reach www.marymount.edu Windows XP Athlon (not 64-bit), VIa 6103: www.marymount.edu comes up in fraction of a second All these are behind a Netgear RP614v2 switch/router -- John Perry -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
This has me scratching my head. After a clean install of 10.2, I am now unable to access my work website (www.marymount.edu). All other websites load up just fine. That IS a strange problem. I also could not connect. I even tried konqueror with several user agent strings. According to Netcraft they are running apache 1.3.33 on Unix, not sure what is causing this one. Wireshark (ethereal) didn't help me much either. That does seem to be
Chip Cooper wrote: the only site so far that hasn't worked. Weird. -- Joe Morris Registered Linux user 231871 -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On 12/17/06, Chip Cooper
This has me scratching my head. After a clean install of 10.2, I am now unable to access my work website (www.marymount.edu). All other websites load up just fine. At first, I thought it was a problem with this specific website, but
really weird ... I can get to it from suse10 using ff or konq -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On 12/18/06, Peter Van Lone
On 12/17/06, Chip Cooper
wrote: This has me scratching my head. After a clean install of 10.2, I am now unable to access my work website (www.marymount.edu). All other websites load up just fine. At first, I thought it was a problem with this specific website, but
really weird ... I can get to it from suse10 using ff or konq -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
No problem here: Old IBM Aptiva PC with OpenSuSE 10.2 just installed, FF and Konqueror both opened the site OK. -- Mark Goldstein -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On Monday 18 December 2006 12:47, Mark Goldstein wrote:
No problem here: Old IBM Aptiva PC with OpenSuSE 10.2 just installed, FF and Konqueror both opened the site OK.
I can't get to www.marymount.edu either. For that matter I can't get to www.keh.com. I just figured something was wrong with the net since everything else works. OTOH both work from my old 9.2 box. Both boxes connect to the same router. Nick -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Nick Zentena wrote:
On Monday 18 December 2006 12:47, Mark Goldstein wrote:
No problem here: Old IBM Aptiva PC with OpenSuSE 10.2 just installed, FF and Konqueror both opened the site OK.
I can't get to www.marymount.edu either. For that matter I can't get to www.keh.com.
OTOH both work from my old 9.2 box. Both boxes connect to the same router.
Nick
I can't get www.keh.com either. I am jealous of mark--wonder why his 10.2 system can access those sites? I disabled ipv6, but it didnt' solve the problem. Interesting Fedora 6 has same problem. That kernel is 2.6.18-1xxx (I believe), and my present kernel is 2.6.18.2-34 Chip
-- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Chip, On Sunday 17 December 2006 19:59, Chip Cooper wrote:
This has me scratching my head. After a clean install of 10.2, I am now unable to access my work website (www.marymount.edu). All other websites load up just fine. At first, I thought it was a problem with this specific website, but if I dual boot to windows, the site comes up just fine, and if I plug my laptop (currently running suse 10.1) to the Ethernet, the site also loads up just fine. So, I imagine something within 10.2 is keeping it from loading--but not sure where the problem is. I've disabled firewall, ipv6, but still cannot load the site. If I ping the site, I get:
I guess I'm in the "something is odd" camp. I have several computers here, a Linux box running SuSE 10.0, VMware on that SuSE 10.0 system, an Intel iMac, a MacBook Pro and a SuSE Linux 10.2 box. The two linux boxes each have a static addresses on the wild Internet connected via DSL. There's a wireless hub using another static external address and a non-routable range of interior network addresses for the WiFI access point and wired access to a second NIC on the 10.2 box and a TiVO box. The DSL modem connects to the Linux boxes and the WiFi router via a NetGear GS108 gigabit switch (not a simple hub). With all that, only the 10.2 box cannot access http://www.marymount.edu/! And this is not a browser or embedded content issue, because I cannot even retrieve the home page with wget, either. All attempts to connect simply hang indefinitely (until the client application gives up). The failure to connect persists when I disable the firewall entirely. The DNS lookup works fine (not surprising, since that's handled via my ISP's DNS servers, of course). If there are log entries with clues to this failure, I cannot find them. Curious indeed. Randall Schulz -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Hi, On Tuesday 19 December 2006 07:50, Randall R Schulz wrote:
Chip,
...
With all that, only the 10.2 box cannot access http://www.marymount.edu/!
I should have made some note of my hardware: ASUS P5B Deluxe Core 2 Duo 32-bit Linux install % /sbin/lspci | grep -i ether 02:00.0 Ethernet controller: Marvell Technology Group Ltd. Unknown device 4364 (rev 12) 05:04.0 Ethernet controller: Marvell Technology Group Ltd. 88E8001 Gigabit Ethernet Controller (rev 14) So it's not Intel Ethernet NIC that's an issue, it would seem. So I started up Wireshark (nee Ethereal) and tried to spot any problems, and my first hunch, as improbable as it seems, was borne out: A bad checksum!! According to Wireshark, the IP packet that was synthesized to convey the HTTP "GET / HTTP/1.0" message had a bad checksum: Checksum: 0x15a8 [incorrect, should be 0xa4bb (maybe caused by checksum offloading?)] Funky, eh? Since this symptom is occurring on different NICs (and different hardware vendor's NICs), the "checksum offloading?" hypothesis doesn't seem right. In fact, the person who pointed the finger at the 10.2 TCP/IP stack as the culprit appears to have been on the right track. Randall Schulz -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Hello again, I've made several Wireshark captures while attempting to retrieve the main HTML content from http://www.marymount.edu/. Every time I use wget to try to retrieve that page, the generated checksum is 0x15a8, regardless of the fact that the packet content is never exactly the same twice and hence should be producing a different checksum each time. On the other hand, when I use Firefox (note, this is version Firefox 2.0), the generated checksum is different for each request, but still wrong. Naturally, the HTTP-level payload is also different, since at a minimum it includes a client identifying string. However, I noticed that when Firefox drives the transaction, there is a TCP retransmission request not seen when I use wget. These retransmissions do exhibit repeated checksums even though the proper checksum changes. I don't know what could be causing this and why it's (so far) manifested only when trying to access this particular site / address, but clearly the checksum generation is the problem. Perhaps it is in some way associated with the destination IP address or range of addresses? Randall Schulz -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On Tuesday 19 December 2006 12:43, Randall R Schulz wrote:
I don't know what could be causing this and why it's (so far) manifested only when trying to access this particular site / address, but clearly the checksum generation is the problem.
Perhaps it is in some way associated with the destination IP address or range of addresses?
No it's also happening with www.keh.com for me. I haven't noticed it on any other site but considering the number of sites out there it's likely hitting others. Nick -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On Tue, 2006-12-19 at 09:43 -0800, Randall R Schulz wrote:
I don't know what could be causing this and why it's (so far) manifested only when trying to access this particular site / address, but clearly the checksum generation is the problem.
I'm by far a guru here, and still waiting on my boxed 10.2, but I wonder if any of you have tried dial-up (modem0) vs your eth0 interface...? All I can say at this point is it workes here, and is a very interesting problem... -- Tom in NM SuSE 9.3/Evolution 11:21am up 8 days 22:02, 3 users, load average: 0.03, 0.12, 0.14 ==== -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Sigh... Apparently the checksum problem is not the issue. Wireshark is reporting this for many (as far as I can tell, all) packets. None of these (except for the marymount.edu site) exhibit the original symptom. I'm not sure why Wireshark is reporting this error spuriously. Perhaps it has something to do with the checksum offloading. Sorry for the false alarm. Randall Schulz -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On Tuesday 19 December 2006 08:11, Randall R Schulz wrote:
ASUS P5B Deluxe Core 2 Duo 32-bit Linux install
How did That happen? Core 2 is x86_64 by default...
According to Wireshark, the IP packet that was synthesized to convey the HTTP "GET / HTTP/1.0" message had a bad checksum:
Lots of, (dare I say most?) Gigabit nics have offloading turned on by default. I think there is a way to suppress same, but have to dig it out. -- _____________________________________ John Andersen
John, On Tuesday 19 December 2006 14:09, John Andersen wrote:
On Tuesday 19 December 2006 08:11, Randall R Schulz wrote:
ASUS P5B Deluxe Core 2 Duo 32-bit Linux install
How did That happen? Core 2 is x86_64 by default...
Says who? What "default?" And why would I want to incur all that overhead when I have no applications that come close to needing the address space of a 64-bit architecture? % uname -a Linux smiley 2.6.18.2-34-bigsmp #1 SMP Mon Nov 27 11:46:27 UTC 2006 i686 i686 i386 GNU/Linux
According to Wireshark, the IP packet that was synthesized to convey the HTTP "GET / HTTP/1.0" message had a bad checksum:
Lots of, (dare I say most?) Gigabit nics have offloading turned on by default. I think there is a way to suppress same, but have to dig it out.
It's clear that the skge driver using the Marvell 8838001 Gigabit Ethernet Controller is doing offloaded checksumming. It's equally clear that the sky2 driver on the SysKonnect controller is not. From what I could find digging around on the Web, drivers that offer the option of controlling offloaded checksum generation (such as the 3c59x) do so via modprobe (see the 6th line of output): % modinfo -p 3c59x debug:3c59x debug level (0-6) options:3c59x: Bits 0-3: media type, bit 4: bus mastering, bit 9: full duplex global_options:3c59x: same as options, but applies to all NICs if options is unset full_duplex:3c59x full duplex setting(s) (1) global_full_duplex:3c59x: same as full_duplex, but applies to all NICs if full_duplex is unset hw_checksums:3c59x Hardware checksum checking by adapter(s) (0-1) flow_ctrl:3c59x 802.3x flow control usage (PAUSE only) (0-1) enable_wol:3c59x: Turn on Wake-on-LAN for adapter(s) (0-1) global_enable_wol:3c59x: same as enable_wol, but applies to all NICs if enable_wol is unset rx_copybreak:3c59x copy breakpoint for copy-only-tiny-frames max_interrupt_work:3c59x maximum events handled per interrupt compaq_ioaddr:3c59x PCI I/O base address (Compaq BIOS problem workaround) compaq_irq:3c59x PCI IRQ number (Compaq BIOS problem workaround) compaq_device_id:3c59x PCI device ID (Compaq BIOS problem workaround) watchdog:3c59x transmit timeout in milliseconds global_use_mmio:3c59x: same as use_mmio, but applies to all NICs if options is unset use_mmio:3c59x: use memory-mapped PCI I/O resource (0-1) When I use modinfo on the skge and sky2 modules no such options are listed, so it seems they don't offer software control over this capability: % modinfo -p sky2 debug:Debug level (0=none,...,16=all) copybreak:Receive copy threshold disable_msi:Disable Message Signaled Interrupt (MSI) idle_timeout:Watchdog timer for lost interrupts (ms) % modinfo -p skge debug:Debug level (0=none,...,16=all) Randall Schulz -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On Tuesday 19 December 2006 13:31, Randall R Schulz wrote:
John,
On Tuesday 19 December 2006 14:09, John Andersen wrote:
On Tuesday 19 December 2006 08:11, Randall R Schulz wrote:
ASUS P5B Deluxe Core 2 Duo 32-bit Linux install
How did That happen? Core 2 is x86_64 by default...
Says who? What "default?"
Well, Randall, since your past responses suggest you are a big wikipedia fan, check here: http://en.wikipedia.org/wiki/Core_2_duo Its a 65bit processor. It will emulate 32bit. You incur some penalty in installation size (disk space) but since the bulk of an os is moving stuff around in memory you will gain the use of 64bit registers to do this if you install the 64bit kernel. -- _____________________________________ John Andersen
On Tuesday 19 December 2006 13:48, John Andersen wrote:
On Tuesday 19 December 2006 13:31, Randall R Schulz wrote:
John,
On Tuesday 19 December 2006 14:09, John Andersen wrote:
On Tuesday 19 December 2006 08:11, Randall R Schulz wrote:
ASUS P5B Deluxe Core 2 Duo 32-bit Linux install
How did That happen? Core 2 is x86_64 by default...
Says who? What "default?"
Well, Randall, since your past responses suggest you are a big wikipedia fan, check here:
http://en.wikipedia.org/wiki/Core_2_duo
Its a 65bit processor. It will emulate 32bit.
Doh! 64bit of course. Extra bit inserted by wayward digits. -- _____________________________________ John Andersen
On Tuesday 19 December 2006 14:48, John Andersen wrote:
On Tuesday 19 December 2006 13:31, Randall R Schulz wrote:
John,
On Tuesday 19 December 2006 14:09, John Andersen wrote:
On Tuesday 19 December 2006 08:11, Randall R Schulz wrote:
ASUS P5B Deluxe Core 2 Duo 32-bit Linux install
How did That happen? Core 2 is x86_64 by default...
Says who? What "default?"
Well, Randall, since your past responses suggest you are a big wikipedia fan, check here:
Wikipedia is a good thing, yes.
Its a 65bit processor. It will emulate 32bit.
65? The question was not about what it can do, but what it's "default" is.
You incur some penalty in installation size (disk space) but since the bulk of an os is moving stuff around in memory you will gain the use of 64bit registers to do this if you install the 64bit kernel.
Think more about this. The primary bottleneck in modern desktop-style computers is the memory interface. The limiting factor there derives from the cycle time of the RAM and the width of the transfer bus. Using a 64 bit processor does not change those parameters. However, it does mean that all addresses and single-word values are 64-bits wide, and transferring that any number of such 64 bit quantities is going to take twice as long as transferring the same number of 32-bit quantities. If all you're doing is a lot of string copies, you won't find much difference. But if you're doing integer arithmetic in programs written in C or C++ and those programs are compiled for the 64-bit ISA, then you're going to be moving a whole lot of high-order zero bits around. If you don't need a 64-bit architecture, you definitely should not use one. I don't need one. Randall Schulz -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On Tuesday 19 December 2006 14:08, Randall R Schulz wrote:
If you don't need a 64-bit architecture, you definitely should not use one. I don't need one.
so why did you buy one? And now that you've discovered, (but not yet admitted) that you installed the wrong kernel, why not do it over and prove or disprove your elaborate theory? -- _____________________________________ John Andersen
On Tuesday 19 December 2006 15:16, John Andersen wrote:
On Tuesday 19 December 2006 14:08, Randall R Schulz wrote:
If you don't need a 64-bit architecture, you definitely should not use one. I don't need one.
so why did you buy one?
It's the best-performing, currently available Intel dual-core processor. I am developing a CPU-intensive application which specifically benefits from multi-core and / or multi-processor systems.
And now that you've discovered, (but not yet admitted) that you installed the wrong kernel, why not do it over and prove or disprove your elaborate theory?
Why on earth must you be so rude? And how, exactly, is the kernel that's running my system just fine, "wrong?" RRS -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On Tuesday 19 December 2006 14:28, Randall R Schulz wrote:
Why on earth must you be so rude?
Pot/Kettle. But it wasn't my intent.
And how, exactly, is the kernel that's running my system just fine, "wrong?"
Its not the native kernel for that hardware. As for your theory, a quick look at the clock times will reveal that the 64bit registers load in the same number of clocks as the 32. And the 64bit address registers, pointer registers, etc also load in the same number of clocks as the 32 bit registers. So moving stuff around in memory is faster doing so in 64bit chunks. The reason for 64bit processors is NOT simply address space. Drive controllers move data to memory at what ever speed and chunk size they are designed to use, but your access to that data in 32bit chunks that take just as long to load as 64bit chunks is slower because it takes more clocks. Look, I've done this exact same thing, loaded the i386 kernel on a Pentium D (also a 64bit dual well processor, in spite of the Pentium name). Since it was a new server with no mission data installed I re-installed using the correct kernel, and it was significantly faster with the 64bit kernel. I did not bench long running applications, or time long compiles, just running yast, kde, and some database loading. Since it runs at init 3 most of the time the 32bit would have worked, and might have gone un-noticed. But it was faster launching applications in 64bit, and the time to load the sql database from raw files was also faster, 37 minutes compared to 52. -- _____________________________________ John Andersen
On Tuesday 19 December 2006 15:49, John Andersen wrote:
On Tuesday 19 December 2006 14:28, Randall R Schulz wrote:
Why on earth must you be so rude?
Pot/Kettle.
Compounding the offense by accusing me of being rude.
But it wasn't my intent.
What, then?
And how, exactly, is the kernel that's running my system just fine, "wrong?"
Its not the native kernel for that hardware.
The hardware is equally i386 and x86_64. A 32-bit kernel is every bit as "native" as a 64-bit one.
As for your theory, a quick look at the clock times will reveal that the 64bit registers load in the same number of clocks as the 32. And the 64bit address registers, pointer registers, etc also load in the same number of clocks as the 32 bit registers. So moving stuff around in memory is faster doing so in 64bit chunks.
It's not about the processor, OK? It's about how much demand 64-bit everything places on the memory subsystem. Why fetch and store so many 64-bit quantities that don't make any use of the upper 32 bits? It's all a big waste of memory bandwidth.
The reason for 64bit processors is NOT simply address space.
If one has no need for 64 bit anything, it's all a waste.
Drive controllers move data to memory at what ever speed and chunk size they are designed to use, but your access to that data in 32bit chunks that take just as long to load as 64bit chunks is slower because it takes more clocks.
Driver controllers are DMA. The processor is not involved in each individual data word transfered.
Look, I've done this exact same thing, loaded the i386 kernel on a Pentium D (also a 64bit dual well processor, in spite of the Pentium name).
Why would results for an ancient Pentium-D imply anything for a Core 2 Duo, which is an entirely different implementation of the x86 architecture?
Since it was a new server with no mission data installed I re-installed using the correct kernel, and it was significantly faster with the 64bit kernel.
Which leaves us with the original accusation--that I made a 32/64 choice. I just stuck the disc in and installed.
I did not bench long running applications, or time long compiles, just running yast, kde, and some database loading. Since it runs at init 3 most of the time the 32bit would have worked, and might have gone un-noticed. But it was faster launching applications in 64bit, and the time to load the sql database from raw files was also faster, 37 minutes compared to 52.
I don't care about those things. I care about CPU-intensive symbolic computations for which 64-bit words, including every single memory address, are dragging around 32 bits of useless zeroes. OK? RRS -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On 2006-12-19 18:38, Randall R Schulz wrote:
On Tuesday 19 December 2006 15:49, John Andersen wrote:
<snip> <snip> *ahem*, guys....
-- The best way to accelerate a computer running Windows is at 9.81 m/s² -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 The Tuesday 2006-12-19 at 16:38 -0800, Randall R Schulz wrote: ...
It's not about the processor, OK? It's about how much demand 64-bit everything places on the memory subsystem. Why fetch and store so many 64-bit quantities that don't make any use of the upper 32 bits? It's all a big waste of memory bandwidth.
Even with a 64 bit system, you can define 32 bits integers, or 16, or 8, and just load the needed bits, not 64 every time. It's only if you tell the compiler to use 64 bit alignement when that might be true. AFAIK, of course :-) - -- Cheers, Carlos E. R. -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.2 (GNU/Linux) Comment: Made with pgp4pine 1.76 iD8DBQFFiJEatTMYHG2NR9URAkh7AJ0f4zxQ4GYMphZym5o84DhGWstDOgCbBOf9 7/uZOeLd62py787sp5CTYFg= =9Sa1 -----END PGP SIGNATURE----- -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On Tuesday 19 December 2006 15:38, Randall R Schulz wrote:
The reason for 64bit processors is NOT simply address space.
If one has no need for 64 bit anything, it's all a waste.
Not really. Its faster. How can that be a waste?
Drive controllers move data to memory at what ever speed and chunk size they are designed to use, but your access to that data in 32bit chunks that take just as long to load as 64bit chunks is slower because it takes more clocks.
Driver controllers are DMA. The processor is not involved in each individual data word transfered.
True, but not relevant to the point of my paragraph. Moving data in memory in 64 bit chunks is faster than moving in 32 bit chunks.
Look, I've done this exact same thing, loaded the i386 kernel on a Pentium D (also a 64bit dual well processor, in spite of the Pentium name).
Why would results for an ancient Pentium-D imply anything for a Core 2 Duo, which is an entirely different implementation of the x86 architecture?
Penitum-D is not ancient. It was only announced in 2005 and the Presler 915 was released in the 1st quarter of 2006. Its a curret processor with emt64 and it works faster with a x86_64 kernel. http://en.wikipedia.org/wiki/Pentium_D
Which leaves us with the original accusation--that I made a 32/64 choice. I just stuck the disc in and installed.
Just exactly what disk did you stick in? Unless you have the boxed set (and I know you don't), you downloaded EITHER a 32bit OR a 64 bit distro. When downloading you can't get both on one DVD. So you made the choice at download time. The installer did not have the choice of picking the optimal kernel. Given a choice (such as when your boxed set arrives with its double layer disk) yast would choose the x86_64 kernel. Randall, run what you want, its your machine. The only reason I commented was you seemed unaware that the core 2 was a 64bit cpu. If you knew that already why not just say you consciously made a choice to stay with 32bits instead of getting all defensive about it? -- _____________________________________ John Andersen
On 12/17/06, Chip Cooper
This has me scratching my head. After a clean install of 10.2, I am now unable to access my work website (www.marymount.edu). All other websites load up just fine. At first, I thought it was a problem with this specific website, but if I dual boot to windows, the site comes up just fine, and if I plug my laptop (currently running suse 10.1) to the Ethernet, the site also loads up just fine. So, I imagine something within 10.2 is keeping it from loading--but not sure where the problem is. I've disabled firewall, ipv6, but still cannot load the site. If I ping the site, I get:
chip@linux-19nx:~> ping -c 3 marymount.edu http://marymount.edu/ PING marymount.edu http://marymount.edu/ (198.100.0.33 http://198.100.0.33/) 56(84) bytes of data.
--- marymount.edu http://marymount.edu/ ping statistics --- 3 packets transmitted, 0 received, 100% packet loss, time 2001ms
and traceroute give me:
linux-19nx:~ # traceroute marymount.edu http://marymount.edu/ traceroute to marymount.edu http://marymount.edu/ ( 198.100.0.33 http://198.100.0.33/), 30 hops max, 40 byte packets 1 192.168.15.1 http://192.168.15.1/ ( 192.168.15.1 http://192.168.15.1/) 0.744 ms 0.386 ms 0.396 ms 2 ip24-250-96-1.dc.dc.cox.net http://ip24-250-96-1.dc.dc.cox.net/ ( 24.250.96.1 http://24.250.96.1/) 18.913 ms 25.217 ms 23.428 m 3 mrfdaggc01.dc.dc.cox.net http://mrfdaggc01.dc.dc.cox.net/ ( 68.100.1.65 http://68.100.1.65/) 16.584 ms 12.408 ms 13.617 ms 4 ip68-100-0-1.dc.dc.cox.net http://ip68-100-0-1.dc.dc.cox.net/ (68.100.0.1 http://68.100.0.1/) 14.263 ms 9.064 ms 11.708 ms 5 mrfddsrj01gex070004.rd.dc.cox.net http://mrfddsrj01gex070004.rd.dc.cox.net/ (68.100.0.161 http://68.100.0.161/) 13.285 ms 13.319 ms 1 .714 ms 6 68.1.0.251 http://68.1.0.251/ ( 68.1.0.251 http://68.1.0.251/) 16.000 ms 14.546 ms 16.563 ms 7 68.105.31.74 http://68.105.31.74/ (68.105.31.74 http://68.105.31.74/) 16.595 ms 15.889 ms 15.372 ms 8 v3495.mpd03.jfk02.atlas.cogentco.com http://v3495.mpd03.jfk02.atlas.cogentco.com/ ( 154.54.6.45 http://154.54.6.45/) 19.512 ms 20.360 ms * 9 * * t9-4.mpd01.dca01.atlas.cogentco.com http://t9-4.mpd01.dca01.atlas.cogentco.com/ ( 154.54.6.138 http://154.54.6.138/) 21.644 ms 10 * * * 11 * * * 12 * * * 13 * * * 14 * * * 15 * * * 16 * * * 17 * * * 18 * * * 19 * * * 20 * * * 21 * * * 22 * * * 23 * * * 24 * * * 25 * * * 26 * * * 27 * * * 28 * * * 29 * * * 30 * * *
Anyone have an idea what is going on with 10.2 that may be causing this? Every other site I try to load comes up just fine. Would appreciate any suggestions--I need to access this site for my job. Thanks!
Chip -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
I can not reach this site as well from my 10.2 i386 ( 2.6.18.2-34) machine. There may be something else - like misconfigured firewall or router or web server. As the OP (Chip) says it is his work website, maybe he can try to talk with their administrator, and review the firewall logs, etc. Maybe these kernels produce some strange TCP packets or something, and the FW rejects them. They may be strict to the standards, and still to look like bad to some filtering software. Cheers -- Svetoslav Milenov (Sunny) Even the most advanced equipment in the hands of the ignorant is just a pile of scrap. -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Chip Cooper wrote:
This has me scratching my head. After a clean install of 10.2, I am now unable to access my work website (www.marymount.edu). All other websites load up just fine. At first, I thought it was a problem with this specific website, but if I dual boot to windows, the site comes up just fine,
I can't even access the site from here with XP and IE??? Mark -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
Onsdag 20 december 2006 10:17 skrev Mark Hounschell:
Chip Cooper wrote:
This has me scratching my head. After a clean install of 10.2, I am now unable to access my work website (www.marymount.edu). All other websites load up just fine. At first, I thought it was a problem with this specific website, but if I dual boot to windows, the site comes up just fine,
I can't even access the site from here with XP and IE???
Mark
Well, I don't even think it's a client problem...I actually think it's a subtle server problem at marymount... I can't even see the site using IE from our place. We're NOT proxying or anything...just going through our normal router, a cheap Linksys. Now, what if the server side has a subtle bug that only manifests itself in certain, very specific situations?? If one reads through all of this thread, what conclusively excludes this possibility? -- ------------------------------ Med venlig hilsen/Best regards Verner Kjærsgaard ------------------------------ -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On 2006-12-20 03:42, Verner Kjærsgaard wrote:
<snip> Now, what if the server side has a subtle bug that only manifests itself in certain, very specific situations?? If one reads through all of this thread, what conclusively excludes this possibility? Absolutely nothing.
-- The best way to accelerate a computer running Windows is at 9.81 m/s² -- To unsubscribe, e-mail: opensuse+unsubscribe@opensuse.org For additional commands, e-mail: opensuse+help@opensuse.org
On Wednesday 20 December 2006 00:42, Verner Kjærsgaard wrote:
Well, I don't even think it's a client problem...I actually think it's a subtle server problem at marymount...
I can't even see the site using IE from our place.
That is strictly a temporary problem. They (marymount.edu) were down temporarily last night. As far as I know, you are the ONLY person reporting access problems with windows. So check it again. -- _____________________________________ John Andersen
participants (18)
-
Andreas
-
Carlos E. R.
-
Chip Cooper
-
Darryl Gregorash
-
Felix Miata
-
Joe Morris (NTM)
-
John Andersen
-
John E. Perry
-
Kenneth Schneider
-
Mark Goldstein
-
Mark Hounschell
-
Mike McMullin
-
Nick Zentena
-
Peter Van Lone
-
Randall R Schulz
-
Sunny
-
Tom Patton
-
Verner Kjærsgaard