Version: 1.2.0 (using 3.5.7 "release 72.6" , openSUSE 10.3) Compiler: Target: i586-suse-linux OS: Linux (i686) release 2.6.22.17-0.1-default I have a Compaq P4 1500MHz with 3 network cards, A intel 10/100/1000 using e1000 driver, A intel 10/100 using e100 driver and a IBM 4/16 token ring. I use ksysguard to monitor network speed for Tx and Rx for each card in kBytes/sec. This works OK in SuSE 10.0 but in 10.2 and 10.3 the display will show just zero, or altinates zero then with a value that is twice the accuall speed. I have tried changing the update interval but that has not worked and I cannot see any other changes that I can make and the system has been updated to the current level. Thankyou Mark. PS You have done a great job on SuSE.
Created attachment 28279 [details] ksysguars showing a 1-min-load of 20000
Created attachment 28280 [details] my computer is melting at 1500 degrees Note the bottom two panels (average load 15 mins, and temperature). Left and right have exactly the same setting, only that the right panels have 5 pixels per timestep, and the left ones 1 pixel per timestep (i.e. the right are closeups of the left panels).
Sorry - the attachments were meant for bug 164851. Don't know how this happened, I was in the other bug report and clicked on "add attachment" there. Can the attachments be moved? Thanks
Is this still an issue? I remember a bug we fixed recently that cause system with no wifi to have no wired adapter shown and therefore 0 for network speed.
Hi Sabastien, Yes this is still a problem that affects 10.2, 10.3 etc. what was the fix that you did? Thanks Mark -----Original Message----- From: Sebastien Martel <sebastiendevel@gmail.com> To: redwoodx@netscape.net Sent: Sun, Jul 19, 2009 6:44 am Subject: [Bug 161266] Network speed monitor not displays or speed value doubles https://bugs.kde.org/show_bug.cgi?id=161266 Sebastien Martel <sebastiendevel@gmail.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |sebastiendevel@gmail.com --- Comment #4 from Sebastien Martel <sebastiendevel gmail com> 2009-07-18 22:44:55 --- Is this still an issue? I remember a bug we fixed recently that cause system with no wifi to have no wired adapter shown and therefore 0 for network speed.
The issue was if the kernel was not built with wifi or if the wifi module is not loaded, e.g. if no /proc/net/wireless file existed then the normal wired connection would not show up in ksysguardd and would have been 0 in ksysguard. To confirm if this was the problem for you, you can check if a file /proc/net/wireless exist, if it doesn't that is your problem and I believe the fix is already in the 4.3 branch and therefore the 4.3 RC that is now publicly available.
Hi Sabastien, The /proc/net/wireless file exist. So that fix did not work for me. But, my system does not have wireless anyway. What it does have it a 10/100 ethernet, 10/100/1000 ethernet and a 16/4 token ring wired network cards. I have a Compaq P4 1500MHz with 3 network cards, An intel 10/100/1000 using e1000 driver, An intel 10/100 on the motherboard using e100 driver and a IBM 4/16 token ring using the Olympic driver. I use ksysguard to monitor network speed for Tx and Rx for each card in kBytes/sec. This works OK in SuSE 10.0 but in 10.2 and 10.3 the display will show just zero, or alternates zero then with a value that is twice the accusal speed. I have tried changing the update interval but that has not worked and I cannot see any other changes that I can make and the system has been updated to the current level. Also I need to set a ring speed option in the olympic token ring driver. I can do it in SuSE 10 but not SuSE 10.2 or 10.3. Thankyou Mark. -----Original Message----- From: Sebastien Martel <sebastiendevel@gmail.com> To: redwoodx@netscape.net Sent: Mon, Jul 20, 2009 9:42 pm Subject: [Bug 161266] Network speed monitor not displays or speed value doubles https://bugs.kde.org/show_bug.cgi?id=161266 --- Comment #6 from Sebastien Martel <sebastiendevel gmail com> 2009-07-20 13:42:51 --- The issue was if the kernel was not built with wifi or if the wifi module is not loaded, e.g. if no /proc/net/wireless file existed then the normal wired connection would not show up in ksysguardd and would have been 0 in ksysguard. To confirm if this was the problem for you, you can check if a file /proc/net/wireless exist, if it doesn't that is your problem and I believe the fix is already in the 4.3 branch and therefore the 4.3 RC that is now publicly available.
Do you have the possibility to test on a newer kde release, is it still an issue with 4.3 RC?
Hi Sebastien, The application is KDE System Guard 1.2.0 using KDE 3.5.7. on 10.3. OpenSUSE 11.1 is using the same program using KDE 3.5.10.?? (It also has KDE 4 installed) My old samba server is running SuSE 10.0 with KDE System Guard 1.2.0 using KDE 3.4.2 without a problem. Now it seems an old application that runs in KDE 3.x. How do we make it use KDE 4.x? Thanks Mark -----Original Message----- From: Sebastien Martel <sebastiendevel@gmail.com> To: redwoodx@netscape.net Sent: Sat, Jul 25, 2009 10:24 pm Subject: [Bug 161266] Network speed monitor not displays or speed value doubles https://bugs.kde.org/show_bug.cgi?id=161266 --- Comment #8 from Sebastien Martel <sebastiendevel gmail com> 2009-07-25 14:24:41 --- Do you have the possibility to test on a newer kde release, is it still an issue with 4.3 RC?
You can update your kde package in openSuse there is a link here of repositories http://en.opensuse.org/KDE/Repositories or compile just ksysguard/ksysguardd yourself if you feel comfortable with that.
I'm not sure whether this is the same problem but here I have 2 100Mbit ethernet cards (one on the motherboard, one PCI) and I see bogus network speeds (download at least is reported as higher than it is actually) from ksysguard. gnome-system-monitor reports the right speed. I believe ksysguard reports double the real speed because it seems to max out at 240kb/s but I have a 1Mbit connection which typically maxs out a 120kb/s. This is using KDE 4.3.1 on ubuntu karmic (which is still alpha).
(In reply to comment #11) > I'm not sure whether this is the same problem but here I have 2 100Mbit > ethernet cards (one on the motherboard, one PCI) and I see bogus network speeds > (download at least is reported as higher than it is actually) from ksysguard. > gnome-system-monitor reports the right speed. If you hover over the graph, does the tooltip say that it's showing "lo" as well ?
> If you hover over the graph, does the tooltip say that it's showing "lo" as > well ? Oh, right, it's saying: * lo * eth0 * eth1 * ppp0 I think in my case counting lo wouldn't be such a problem probably but the problem is likely counting eth0 and ppp0 because ppp0 is on top of eth0, I'm doing PPPoE to talk to the modem of my ISP, over PPPoE. So if it counts eth0 and ppp0 then it's going to count the traffic twice. So I think that's most likely the explanation for this problem. I don't know how gnome-network-manager gets it right, because AFAIK the ppp interface might be bound to something else than an ethernet card and then it probably should be counted, not discarded. Should I open a new bug?
> So if it counts eth0 and ppp0 then it's going to count the traffic twice. Ah, this is indeed interesting. I hadn't even thought of that. Just to confirm, when you hover over it I think it gives the data rate. Can you see there if eth0 and ppp0 have similar numbers? Anyone know how how to fix this? How can I easily determine that ppp0 and eth0 are the same device?
> Just to confirm, when you hover over it I think it gives the data > rate. Can you see there if eth0 and ppp0 have similar numbers? similar yes, but not identical. like 118kb/s and 120. 114 and 116. 118 and 116. but both numbers taken independently are OK. Summing them is wrong. > Anyone know how how to fix this? How can I easily determine that ppp0 > and eth0 are the same device? I'll try to look at the source of gnome-system-monitor but i don't know if i'll have the time and if i'll find the relevant piece of code. in any case it can't be parsing of any config file or something like that because I have PPPoE currently configured by pppoeconf but it could be also network-manager.
>> Just to confirm, when you hover over it I think it gives the data >> rate. Can you see there if eth0 and ppp0 have similar numbers? > > similar yes, but not identical. like 118kb/s and 120. 114 and 116. 118 and 116. > but both numbers taken independently are OK. Summing them is wrong. Okay then yes, this is the problem >> Anyone know how how to fix this? How can I easily determine that ppp0 >> and eth0 are the same device? > > I'll try to look at the source of gnome-system-monitor but i don't know if i'll > have the time and if i'll find the relevant piece of code. > in any case it can't be parsing of any config file or something like that > because I have PPPoE currently configured by pppoeconf but it could be also > network-manager. Thanks. I hope that they don't just take the largest value, instead of summing. John
I'm not so sure but I may have found the way they do it. See line 489 in: http://git.gnome.org/cgit/gnome-system-monitor/tree/src/load-graph.cpp /* Skip interfaces without any IPv4/IPv6 address (or those with only a LINK ipv6 addr) However we need to be able to exclude these while still keeping the value so when they get online (with NetworkManager for example) we don't get a suddent peak. Once we're able to get this, ignoring down interfaces will be possible too. */ if (not (netload.flags & (1 << GLIBTOP_NETLOAD_ADDRESS6) and netload.scope6 != GLIBTOP_IF_IN6_SCOPE_LINK) and not (netload.flags & (1 << GLIBTOP_NETLOAD_ADDRESS))) continue; Now my ppp0 does have an IP but my eth0 does not have an IP. So if I read this right, that would explain how gnome-system-monitor does it. I guess he would count only ppp0 and throws away eth0 because eth0 doesn't have an IP. Right now I am on internet and here are the relevant parts of my ifconfig: eth0 Link encap:Ethernet HWaddr 00:0e:a6:4c:45:73 inet6 addr: fe80::20e:a6ff:fe4c:4573/64 Scope:Link UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:21746 errors:0 dropped:0 overruns:0 frame:5 TX packets:16077 errors:0 dropped:0 overruns:0 carrier:0 collisions:56 txqueuelen:1000 RX bytes:29079380 (29.0 MB) TX bytes:1870680 (1.8 MB) Interrupt:23 Base address:0x8000 (no IP!) ppp0 Link encap:Point-to-Point Protocol inet addr:86.61.21.172 P-t-P:213.250.19.90 Mask:255.255.255.255 UP POINTOPOINT RUNNING NOARP MULTICAST MTU:1492 Metric:1 RX packets:21646 errors:0 dropped:0 overruns:0 frame:0 TX packets:15997 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:3 RX bytes:28596815 (28.5 MB) TX bytes:1515998 (1.5 MB) (there is an IP!)
Ah, if only all users reported such great bug reports.
Dear Bug Submitter, This bug has been stagnant for a long time. Could you help us out and re-test if the bug is valid in the latest version? I am setting the status to NEEDSINFO pending your response, please change the Status back to REPORTED when you respond. Thank you for helping us make KDE software even better for everyone!
Dear Bug Submitter, This is a reminder that this bug has been stagnant for a long time. Could you help us out and re-test if the bug is valid in the latest version? Thank you for helping us make KDE software even better for everyone!
ksysguard is no longer maintained, in Plasma 6 there is the Plasma system monitor for this task. If your issue still happens with the Plasma 6 replacement, please re-open and we can move this bug to the new product, thanks!