Re: [Samba] Streaming large videos causes server's networking to crash
Well, I figured it out. Turns out the problem was actually the Linux kernel itself. The bug report is here: https://bugs.launchpad.net/ubuntu/+source/linux/+bug/347711 Just as the comments suggested, changing to the mainline kernel has apparently solved the problem. I've been playing video for 18 hours solid now and everything is fine. Thanks for your assistance guys. -- Paul A. Quinn Fissler wrote: Your tsk tsk etc No - not at all - you inferred that incorrectly. My point was that I'd not bothered to read your email, so asked you a question I didn't need to ask. -- To unsubscribe from this list go to the following URL and read the instructions: https://lists.samba.org/mailman/options/samba
[Samba] Streaming large videos causes server's networking to crash
Greetings all; I really hope someone can help me. I recently upgraded to the most recent version (from a year-old version) of Samba, on a server now running on Ubuntu Server Jaunty Jackalope x64. After upgrading, whenever I play large media files on my Vista machine streamed from the Samba server, the networking functionality on the server drops out. Sometimes it happens after playing a video for 2 minutes, sometimes it takes 2 hours. After it drops out, running /etc/init.d/networking restart on the local terminal restores networking functionality. Rebooting also has this effect. This never happened before I upgraded, so I doubt it's a hardware problem. I see the following in the log for the computer the video was being viewed on: [2009/07/29 01:13:21, 0] lib/util_sock.c:write_data(1136) [2009/07/29 01:13:21, 0] lib/util_sock.c:get_peer_addr_internal(1676) getpeername failed. Error was Transport endpoint is not connected write_data: write failure in writing to client 0.0.0.0. Error Connection reset by peer [2009/07/29 01:13:21, 0] smbd/process.c:srv_send_smb(74) Error writing 4159 bytes to client. -1. (Transport endpoint is not connected) [2009/07/29 01:13:21, 1] smbd/service.c:close_cnum(1323) murahime (:::192.168.1.100) closed connection to service Paul's Folder and the following in log.nmbd: [2009/07/29 01:13:23, 0] nmbd/nmbd_become_lmb.c:become_local_master_stage2(395) * Samba name server DITE is now a local master browser for workgroup WORKGROUP on subnet 192.168.1.200 * Any ideas? --Paul Accisano -- To unsubscribe from this list go to the following URL and read the instructions: https://lists.samba.org/mailman/options/samba
Re: [Samba] Streaming large videos causes server's networking to crash
Sure enough, I apparently have a Realtek device as well. But swapping motherboards is absolutely not an option for me... Surely there's something that can be done! Here's the output of lshw -C network: *-network description: Ethernet interface product: RTL8111/8168B PCI Express Gigabit Ethernet controller vendor: Realtek Semiconductor Co., Ltd. physical id: 0 bus info: p...@:05:00.0 logical name: eth0 version: 01 serial: 00:1d:92:60:6d:b8 size: 1GB/s capacity: 1GB/s width: 64 bits clock: 33MHz capabilities: pm vpd msi pciexpress bus_master cap_list ethernet physical tp mii 10bt 10bt-fd 100bt 100bt-fd 1000bt 1000bt-fd autonegotiation configuration: autonegotiation=on broadcast=yes driver=r8169 driverversion=2.3LK-NAPI duplex=full ip=192.168.1.200 latency=0 link=yes module=r8169 multicast=yes port=MII speed=1GB/s This is from the contents of /var/log/kern.log at the time of the last crash: Jul 29 01:13:21 DITE kernel: [ 4616.040039] [ cut here ] Jul 29 01:13:21 DITE kernel: [ 4616.040044] WARNING: at /build/buildd/linux-2.6.28/net/sched/sch_generic.c:226 dev_watchdog+0x270/0x280() Jul 29 01:13:21 DITE kernel: [ 4616.040048] NETDEV WATCHDOG: eth0 (r8169): transmit timed out Jul 29 01:13:21 DITE kernel: [ 4616.040050] Modules linked in: video output input_polldev lp ppdev pcspkr snd_hda_intel iTCO_wdt iTCO_vendor_support psmouse serio_raw snd_pcm parport_pc snd_timer parport snd intel_agp soundcore snd_page_alloc shpchp r8169 mii arcmsr fbcon tileblit font bitblit softcursor Jul 29 01:13:21 DITE kernel: [ 4616.040080] Pid: 0, comm: swapper Not tainted 2.6.28-11-server #42-Ubuntu Jul 29 01:13:21 DITE kernel: [ 4616.040083] Call Trace: Jul 29 01:13:21 DITE kernel: [ 4616.040085] IRQ [80250927] warn_slowpath+0xb7/0xf0 Jul 29 01:13:21 DITE kernel: [ 4616.040098] [80602817] ? __udp_queue_rcv_skb+0x27/0xf0 Jul 29 01:13:21 DITE kernel: [ 4616.040103] [80604c7b] ? udp_queue_rcv_skb+0x1ab/0x350 Jul 29 01:13:21 DITE kernel: [ 4616.040108] [80604f57] ? __udp4_lib_mcast_deliver+0x137/0x250 Jul 29 01:13:21 DITE kernel: [ 4616.040114] [802199e6] ? read_tsc+0x16/0x40 Jul 29 01:13:21 DITE kernel: [ 4616.040120] [802708b9] ? getnstimeofday+0x59/0xe0 Jul 29 01:13:21 DITE kernel: [ 4616.040125] [8041cd0a] ? strlcpy+0x4a/0x60 Jul 29 01:13:21 DITE kernel: [ 4616.040129] [805caf70] dev_watchdog+0x270/0x280 Jul 29 01:13:21 DITE kernel: [ 4616.040133] [8026e61c] ? sched_clock_cpu+0xcc/0x160 Jul 29 01:13:21 DITE kernel: [ 4616.040138] [802199e6] ? read_tsc+0x16/0x40 Jul 29 01:13:21 DITE kernel: [ 4616.040141] [805cad00] ? dev_watchdog+0x0/0x280 Jul 29 01:13:21 DITE kernel: [ 4616.040146] [8025bdf9] run_timer_softirq+0x179/0x260 Jul 29 01:13:21 DITE kernel: [ 4616.040152] [802736bf] ? clockevents_program_event+0x4f/0x90 Jul 29 01:13:21 DITE kernel: [ 4616.040156] [80256a5c] __do_softirq+0x9c/0x170 Jul 29 01:13:21 DITE kernel: [ 4616.040161] [80213d8c] call_softirq+0x1c/0x30 Jul 29 01:13:21 DITE kernel: [ 4616.040165] [80214ffd] do_softirq+0x5d/0xa0 Jul 29 01:13:21 DITE kernel: [ 4616.040169] [802567dd] irq_exit+0x8d/0xa0 Jul 29 01:13:21 DITE kernel: [ 4616.040174] [80227658] smp_apic_timer_interrupt+0x88/0xc0 Jul 29 01:13:21 DITE kernel: [ 4616.040179] [80213668] apic_timer_interrupt+0x88/0x90 Jul 29 01:13:21 DITE kernel: [ 4616.040181] EOI [8021a95a] ? mwait_idle+0x4a/0x50 Jul 29 01:13:21 DITE kernel: [ 4616.040190] [80210dd2] ? enter_idle+0x22/0x30 Jul 29 01:13:21 DITE kernel: [ 4616.040194] [80210e85] ? cpu_idle+0x65/0xc0 Jul 29 01:13:21 DITE kernel: [ 4616.040200] [80689d3c] ? rest_init+0x5c/0x70 Jul 29 01:13:21 DITE kernel: [ 4616.040203] ---[ end trace 07c323730455bb85 ]--- Jul 29 01:13:21 DITE kernel: [ 4616.080950] r8169: eth0: link up Quinn Fissler wrote: Which network device are you using? when it happens, do a dmesg|tail on the server. Anything interesting there? I had a similar symptoms which turned out to be caused by the Realtek device on my motherboard and the driver in my Linux distro. (and when I fixed that by using another driver, my machine was still not stable so I swapped the motherboard out for an older one - that's when I found I also had a faulty sata data cable... Just be warned that problems don't always come alone) 2009/7/29 Paul Accisano shiningmasam...@gmail.com mailto:shiningmasam...@gmail.com Greetings all; I really hope someone can help me. I recently upgraded to the most recent version (from a year-old version) of Samba, on a server now running on Ubuntu Server Jaunty Jackalope x64. After upgrading, whenever I play large media files on my Vista machine streamed from the Samba server
Re: [Samba] Streaming large videos causes server's networking to crash
Of course. The only problem is that the PCI-E 16x slot next to my PCI-E 1x slot is taken with a RAID controller, so my options are limited to tiny cards. http://www.newegg.com/Product/Product.aspx?Item=N82E16833166015 That one looks fine, but the Newegg comments seem to indicate it is also a Realtek-based device. Would it not have the same problem? To Quinn, here's the output of lsmod: Module Size Used by joydev 20864 0 hid_microsoft 12676 0 usbhid 47040 0 video 29204 0 output 11648 1 video input_polldev 12688 0 lp 19588 0 ppdev 16904 0 pcspkr 11136 0 snd_hda_intel 557364 0 iTCO_wdt 21712 0 iTCO_vendor_support12420 1 iTCO_wdt psmouse64028 0 serio_raw 14468 0 snd_pcm99336 1 snd_hda_intel parport_pc 45096 1 snd_timer 34064 1 snd_pcm parport49584 3 lp,ppdev,parport_pc snd78792 3 snd_hda_intel,snd_pcm,snd_timer intel_agp 39280 0 soundcore 16800 1 snd snd_page_alloc 18704 2 snd_hda_intel,snd_pcm shpchp 44572 0 r8169 46596 0 mii14464 1 r8169 arcmsr 36096 3 fbcon 49792 0 tileblit 11264 1 fbcon font 17024 1 fbcon bitblit14464 1 fbcon softcursor 10368 1 bitblit --Paul A. Eero Volotinen wrote: Paul Accisano kirjoitti: Sure enough, I apparently have a Realtek device as well. But swapping motherboards is absolutely not an option for me... Surely there's something that can be done! Well, can you buy better ethernet card? Jul 29 01:13:21 DITE kernel: [ 4616.040048] NETDEV WATCHDOG: eth0 (r8169): transmit timed out I think this is related to ethernet controller or irq issues on motherboard. -- Eero -- To unsubscribe from this list go to the following URL and read the instructions: https://lists.samba.org/mailman/options/samba
Re: [Samba] Streaming large videos causes server's networking to crash
That link certainly sounds promising. I'll try to induce another crash after trying the things you specified and see what happens. I really appreciate your helping me even though this is looking like it's not even Samba-related. Your tsk tsk implies that posting one of those logs was a security risk. For the sake of learning, which log should I not have posted? --Paul A. Quinn Fissler wrote: It could be that tweaking your driver parameters with your curent card could cure the problem... We know it's the r8169 from your other logs (which I didn't read - tsk tsk) Look at your irq allocations cat /proc/interrupts See the network driver - note its IRQ - have a look at its corresponding stats in /proc/irq/you noted/files - there are counters there - maybe they'll help diagnose (unhandled interrupts etc). Look at using an alternative driver... I just googled and found this: http://www.jamesonwilliams.com/hardy-r8168 2009/7/29 Paul Accisano shiningmasam...@gmail.com mailto:shiningmasam...@gmail.com Of course. The only problem is that the PCI-E 16x slot next to my PCI-E 1x slot is taken with a RAID controller, so my options are limited to tiny cards. http://www.newegg.com/Product/Product.aspx?Item=N82E16833166015 That one looks fine, but the Newegg comments seem to indicate it is also a Realtek-based device. Would it not have the same problem? To Quinn, here's the output of lsmod: Module Size Used by joydev 20864 0 hid_microsoft 12676 0 usbhid 47040 0 video 29204 0 output 11648 1 video input_polldev 12688 0 lp 19588 0 ppdev 16904 0 pcspkr 11136 0 snd_hda_intel 557364 0 iTCO_wdt 21712 0 iTCO_vendor_support12420 1 iTCO_wdt psmouse64028 0 serio_raw 14468 0 snd_pcm99336 1 snd_hda_intel parport_pc 45096 1 snd_timer 34064 1 snd_pcm parport49584 3 lp,ppdev,parport_pc snd78792 3 snd_hda_intel,snd_pcm,snd_timer intel_agp 39280 0 soundcore 16800 1 snd snd_page_alloc 18704 2 snd_hda_intel,snd_pcm shpchp 44572 0 r8169 46596 0 mii14464 1 r8169 arcmsr 36096 3 fbcon 49792 0 tileblit 11264 1 fbcon font 17024 1 fbcon bitblit14464 1 fbcon softcursor 10368 1 bitblit --Paul A. Eero Volotinen wrote: Paul Accisano kirjoitti: Sure enough, I apparently have a Realtek device as well. But swapping motherboards is absolutely not an option for me... Surely there's something that can be done! Well, can you buy better ethernet card? Jul 29 01:13:21 DITE kernel: [ 4616.040048] NETDEV WATCHDOG: eth0 (r8169): transmit timed out I think this is related to ethernet controller or irq issues on motherboard. -- Eero -- To unsubscribe from this list go to the following URL and read the instructions: https://lists.samba.org/mailman/options/samba
