Network problems and hangs... losing faith... (solved)


114 posts   Page 4 of 5   1, 2, 3, 4, 5
by mahjongg » Tue Aug 07, 2012 11:30 pm
You seem to be implying that directly connected low current high speed USB devices cause problems with the Pi. That's news to me.

News for me too, I wasn't aware I was implying that. Sorry Englisch is my second language, I might make mistakes. (did I?)
User avatar
Forum Moderator
Forum Moderator
Posts: 5738
Joined: Sun Mar 11, 2012 12:19 am
by fredjam » Wed Aug 08, 2012 2:09 am
mahjongg wrote:
lostintime wrote:I don't expect the Pi to work "plug and play with any bit of hardware you choose". A reasonable expectation in my book is for the Pi to work perfectly with USB hubs. It doesn't.

not to be pedantic, only to further knowledge, but you do understand that the PI has a complete hub inside?! One that has three ports, two of them are routed out to connectors, the third is connected to a built in USB Ethernet adapter! So if you say "I expect the PI to work with a HUB", you in fact are expecting the bare PI to work (with its built in hub).

If you have an issue with connectivity with a further hub you you should rephrase your statement to "I expect the PI to work with a hub behind its existing hub".

Again, I'm not trying to be pedantic, I'm just trying to ensure that everyone understands the issues, so that a solution can be found. It's no good to try to understand the problem if you are not aware of all the facts.

So do say, the PI works perfectly with its built in hub, and standard low current low speed USB devices connected to its built in hub, there are problems through if you try to connect another hub to the one built in.

But maybe I'm mistaken, and there are really problems with the PI's built in hub, connected to a mixture of USB devices.

Note, THE PI IS ONLY COMPATIBLE WITH LOW CURRENT LOW SPEED DEVICES CONNECTED TO ITS HUB PORTS, I apologise for shouting!


Well you should apologise for shouting. If the pi will only work in the way you suggest then
it will not be good enough for the purpose it was designed for. There is a problem with
the USB code and I expect that it will be fixed and that the pi will work with a greater
range of external hubs. At present even 'standard' low current usb keyboards and mice
plugged directly into the pi's usb ports can cause problems. You would know that if you had
bothered to read more of the posts in these forums. and had looked at the lists of working and
non-working peripherals.

I suggest you also look at the definition of 'plug and play', it doesn't mean plug and
might play.
Posts: 83
Joined: Thu Jul 19, 2012 3:19 am
Location: London UK
by rspitz » Wed Aug 08, 2012 8:43 am
It seems to me that the hype the RaspberryPi has stirred up in the worldwide geek community is cooling down, as more and more people are finally getting their Pis delivered and are now running into the same USB problems. But I sincerely hope that this increases the chances of getting the problem fixed, as more geeks are working on this. However, there is still a risk that it is a fundamental hardware issue, then we'd have to write this off as an instructive experience :(
As a side note: I cannot subscribe to the opinion that the RaspberryPi is cheap. Take into consideration that it is just the naked board and you need to count in at least the cost of a case, a power supply and a powered USB hub (the 140mA polyfuse limit on the on-board ports is a joke), then you get into regions where other low-cost ARM devices can compete.
Posts: 55
Joined: Tue Jul 31, 2012 7:25 pm
by mahjongg » Thu Aug 09, 2012 6:23 pm
mahjongg wrote:
You seem to be implying that directly connected low current high speed USB devices cause problems with the Pi. That's news to me.

News for me too, I wasn't aware I was implying that. Sorry English is my second language, I might make mistakes. (did I?)

What I was actually trying to explain is that due to the voltage drop caused by the polyfuses only devices that are designed following the USB consortium's specification for "single current units devices" are compatible. Perhaps calling these devices "low current low speed" devices as a simplification was a bit misleading, because the polyfuses don't impose the need to use only USB 1.0 compatible, ("low speed" devices) they determine only only what their current consumption, and minimum working voltage should be.

the vast majority of all USB keyboards comply to the "single current units specification", and virtually all mice. One reason for keyboards not to work is if they use too much current (by including a hub, trackpad or lighted keys) or are in fact PS/2 keyboards used with an adapter.

So in fact the PI is compatible for its intended purpose, that is "to supply kids with a working computer, (including a keyboard and mouse)".
Beyond that, as the FAQ puts it.
Model A has one USB port and Model B has 2. Beyond this, mice, keyboards, network adapters and external storage will all connect via a USB hub.

As there are only two USB ports on a PI, a third device logically requires a hub, and as the PI doesn't provide enough power for such a third USB device (only 200mA allocated for two 100mA USB devices) this should be a powered hub.

In conclusion, high speed and high current USB devices do work on the PI, but high current devices must be powered by a hub, and high speed devices work directly on a PI only when they are also "single current unit" devices.

Perhaps there are some software kinks to solve in the software, but they are being worked on.
User avatar
Forum Moderator
Forum Moderator
Posts: 5738
Joined: Sun Mar 11, 2012 12:19 am
by halfabiscuit » Thu Aug 09, 2012 7:18 pm
Well first off let me say respect to RS Components for the RMA process. I posted it by recorded delivery Tuesday pm and they dispatched a replacement Wednesday pm which arrived today and (praise the lord) I have a working pi.
I think my sanity has been restored - spent several months trying to get the thing working and with this replacement I had it happily headless and working on my network within about 10 minutes of opening the box.
So I guess it must have been a hardware issue as nothing else in my setup has changed. My advice to anyone else getting the same symptoms as me is to RMA it ASAP.

Now let the fun begin!
Posts: 26
Joined: Thu Jun 21, 2012 10:55 am
by mahjongg » Fri Aug 10, 2012 1:51 am
well, exactly. You can have much talk about this and that, and that when an individual user cannot get his ethernet to work, "there must be a fundamental problem with the design", but in the end it was all based on false assumptions. Yes, sometimes (for any electronic device it would be deemed quite good if it would happen less than 0.5 % off all produced devices) a device is simply broken.
Even though (as I was told) all PI's leave the factory tested on all mayor points.

I'm not saying there might not be some software issues with the USB stack, but they are not so consequential that they affect most people most of the time, only some people some of the time.
And they are software issues, so they can, and will be, resolved.
User avatar
Forum Moderator
Forum Moderator
Posts: 5738
Joined: Sun Mar 11, 2012 12:19 am
by halfabiscuit » Fri Aug 10, 2012 8:05 am
I think the problem is that the symptoms I experienced due to faulty hardware are similar to what other people have experienced due to USB / power issues. I really did the RMA as a last resort and I half expected RS to send it back saying it was fine.
Maybe the whole process could have been smoother if the pi was sold as a boxed product like your old home computer with a power supply, SD card, keyboard etc. that had all been tested and proved 100% to work. Then if people had problems with stability they could be fairly sure that the problem was hardware related and could RMA it. They could add a disclaimer saying "this product has been verified to work with these peripherals - you can use other peripherals but you do so at your own risk".
Hindsight is a wonderful thing though and I imagine most people on this forum would baulk at such a walled garden approach. But if I was your average teenager trying to learn to program on the pi I would have given up a long time ago after my experiences.
Posts: 26
Joined: Thu Jun 21, 2012 10:55 am
by mahjongg » Fri Aug 10, 2012 11:52 am
Well, AFAIK all PI's are fully tested when they leave the factory, but that doesn't rule out DOA's completely, also you have to add your own SD-card, which is like selling PC's without hard-disks.

I fully expect the situation to stabilize when most of the more notorious bugs are ironed out, and it becomes more well know what the PI works with an not, and why.
User avatar
Forum Moderator
Forum Moderator
Posts: 5738
Joined: Sun Mar 11, 2012 12:19 am
by obcd » Fri Aug 10, 2012 12:16 pm
They don't sell it all boxed, but they do sell a compatible power supply and a sd card with a preinstalled OS. They even already sell the housings for it. The problem is I am like most people thinking, I have a 5V adapter from my cellphone I can use. And why shouldn't I use that sd card from my digital camera. I still have an usb keyboard and mouse as well...If things don't work with them, it's often difficult to pinpoint the issue, specially since indeed a lot of factors lead to the same bad working behavour. This also explains why the question keeps repeating: "what's your voltage between TP1 and TP2..It's flexibility is it's strength and it's weak point at the same time. How many different screens can you connect....? How many different brand and type of mice... How many keyboards...
This makes it all great for a guy like me who wan't trow away anything. It make it fustrating for a beginner who isn't getting a login screen. It also makes it difficult for helpfull forum members to find the cause of the issue.
Posts: 890
Joined: Sun Jul 29, 2012 9:06 pm
by DrK » Mon Aug 13, 2012 1:04 pm
Had similar problems - bought Pi from Farnell, hub, keyboard and mouse (wireless) from ModMyPi -all was well when mouse/keyboard were attached directly to the Pi, wired network was fine - however, move the mouse or keyboard to the (powered) hub and the network dropped out. I noted that the pi tried to start when the only power to the board was via the hub (of course it failed) and read everyone comments about power draw, so I bravely (for me!) opened the hub, and using a circular saw bit in a Dremel tool, cut through the power connector on the board between the power input and the mini-b USB board that connects to the pi, so now the hub used power ONLY from its own supply, it COULD NOT draw from the pi. BUT the problem remained and the solution was cheap and simple - i replaced the mouse (Logitech M195) with a wired mouse (Trust, no code number, £2 in Morrisons!) and now all is well.... (I also replaced the Xenta Chiclet wireless keyboard with a Trust £4 wired keyboard to "be safe") - now the hub can support them and a USB stick with wired network fully functional.

Anyone need a nearly-new wireless mouse?

Ken
Posts: 2
Joined: Mon Aug 13, 2012 12:50 pm
by ggodart » Thu Aug 30, 2012 9:15 pm
Not sure if this is related, but I'm setting up my PI and installing stuff with apt-get, this fails when connecting to the mirrordirector, after which the network goes down and I cannot even ping anything, network restart has no effect and I have to reboot.
I can workaround this by running a ping session in another window in parallel to and before the apt-get, I suspect this is a software issue and probably a timeout. I'm in the UK and connects to the mirrordirector are really variable - ping takes between 50ms and 3,000ms.
Posts: 6
Joined: Thu Aug 30, 2012 9:02 pm
by f-unix » Sun Sep 23, 2012 5:56 pm
Hi everybody,

I have the same infamous issue on my PI.
Setup: external powered hub connected with second nic and keyboard on it.
The PI is acting as router/firewall for the internal network.
After random time I got:

Code: Select all
Sep 23 13:26:41 raspberrypi kernel: [ 3652.382405] ------------[ cut here ]------------
Sep 23 13:26:41 raspberrypi kernel: [ 3652.382468] WARNING: at net/sched/sch_generic.c:255 dev_watchdog+0x2a0/0x2c0()
Sep 23 13:26:41 raspberrypi kernel: [ 3652.382486] NETDEV WATCHDOG: eth1 (MOSCHIP usb-ethernet driver): transmit queue 0 timed out
Sep 23 13:26:41 raspberrypi kernel: [ 3652.382500] Modules linked in: xt_tcpudp xt_state xt_conntrack iptable_filter ipt_MASQUERADE iptable_nat nf_nat nf_conntrack_ipv4 nf_conntrack nf_defrag_ipv4 ip_tables x_tables snd_bcm2835 snd_pcm snd_seq snd_timer snd_seq_device snd snd_page_alloc hid_logitech_dj mcs7830
Sep 23 13:26:41 raspberrypi kernel: [ 3652.382639] [<c0013e1c>] (unwind_backtrace+0x0/0xf0) from [<c0028e78>] (warn_slowpath_common+0x4c/0x64)
Sep 23 13:26:41 raspberrypi kernel: [ 3652.382675] [<c0028e78>] (warn_slowpath_common+0x4c/0x64) from [<c0028f24>] (warn_slowpath_fmt+0x30/0x40)
Sep 23 13:26:41 raspberrypi kernel: [ 3652.382708] [<c0028f24>] (warn_slowpath_fmt+0x30/0x40) from [<c02eb704>] (dev_watchdog+0x2a0/0x2c0)
Sep 23 13:26:41 raspberrypi kernel: [ 3652.382758] [<c02eb704>] (dev_watchdog+0x2a0/0x2c0) from [<c003433c>] (run_timer_softirq+0x114/0x2a4)
Sep 23 13:26:41 raspberrypi kernel: [ 3652.382793] [<c003433c>] (run_timer_softirq+0x114/0x2a4) from [<c002e980>] (__do_softirq+0x90/0x128)
Sep 23 13:26:41 raspberrypi kernel: [ 3652.382824] [<c002e980>] (__do_softirq+0x90/0x128) from [<c002ee40>] (irq_exit+0x90/0x98)
Sep 23 13:26:41 raspberrypi kernel: [ 3652.382862] [<c002ee40>] (irq_exit+0x90/0x98) from [<c000e880>] (handle_IRQ+0x34/0x84)
Sep 23 13:26:41 raspberrypi kernel: [ 3652.382898] [<c000e880>] (handle_IRQ+0x34/0x84) from [<c037ccf4>] (__irq_svc+0x34/0xc8)
Sep 23 13:26:41 raspberrypi kernel: [ 3652.382932] [<c037ccf4>] (__irq_svc+0x34/0xc8) from [<c000e990>] (default_idle+0x24/0x28)
Sep 23 13:26:41 raspberrypi kernel: [ 3652.382961] [<c000e990>] (default_idle+0x24/0x28) from [<c000eb78>] (cpu_idle+0x9c/0xc4)
Sep 23 13:26:41 raspberrypi kernel: [ 3652.383001] [<c000eb78>] (cpu_idle+0x9c/0xc4) from [<c04c2718>] (start_kernel+0x280/0x2c8)
Sep 23 13:26:41 raspberrypi kernel: [ 3652.383020] ---[ end trace 10d00bc51283e5e1 ]---


Only way to solve is to reboot.
I'm not quite sure if I understood correctly what is the reason of the issue. Is it something that happens on faulty boards ? In this case is there a warranty to ensure a replacement ?

Thanks.
Posts: 1
Joined: Sun Sep 23, 2012 5:45 pm
by jesusdf » Thu Nov 01, 2012 9:12 pm
Hi! I'm experiencing the exactly same sympthoms: 3 sd cards, a lot of PSUs (including a MOLEX2USB home made to use a computer PSU) and after around 250 ping packets if just stops responding.

I've made many tests with the power supply, and different cables until I got 4.90v inside the raspberry (the psu outputs 5.40v, the cable loses 0.20v aprox.).

Kern.log and messages seem fine, no reboots, this is what I get with the latest 2012-10-28 raspbian image just written to a 2 GB sd, never logged in, only power cable and network plugged in, and pinging from another computer (DHCP works...):

Oct 28 22:59:40 raspberrypi kernel: imklog 5.8.11, log source = /proc/kmsg started.
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] Initializing cgroup subsys cpu
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] Linux version 3.2.27+ (dc4@dc4-arm-01) (gcc version 4.7.2 20120731 (prerelease) (crosstool-NG linaro-1.13.1+bzr2458 - Linaro GCC 2012.08) ) #250 PREEMPT Thu Oct 18 19:03:02 BST 2012
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] CPU: ARMv6-compatible processor [410fb767] revision 7 (ARMv7), cr=00c5387d
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] CPU: PIPT / VIPT nonaliasing data cache, VIPT nonaliasing instruction cache
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] Machine: BCM2708
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] Memory policy: ECC disabled, Data cache writeback
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] On node 0 totalpages: 49152
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] free_area_init_node: node 0, pgdat c050db28, node_mem_map c05b6000
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] Normal zone: 384 pages used for memmap
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] Normal zone: 0 pages reserved
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] Normal zone: 48768 pages, LIFO batch:15
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] pcpu-alloc: s0 r0 d32768 u32768 alloc=1*32768
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] pcpu-alloc: [0] 0
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] Built 1 zonelists in Zone order, mobility grouping on. Total pages: 48768
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] Kernel command line: dma.dmachans=0x7f35 bcm2708_fb.fbwidth=656 bcm2708_fb.fbheight=416 bcm2708.boardrev=0x2 bcm2708.serial=0x88a938eb smsc95xx.macaddr=B8:27:EB:A9:38:EB sdhci-bcm2708.emmc_clock_freq=100000000 vc_mem.mem_base=0xc000000 vc_mem.mem_size=0x10000000 dwc_otg.lpm_enable=0 console=ttyAMA0,115200 kgdboc=ttyAMA0,115200 console=tty1 root=/dev/mmcblk0p2 rootfstype=ext4 elevator=deadline rootwait
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] PID hash table entries: 1024 (order: 0, 4096 bytes)
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] Dentry cache hash table entries: 32768 (order: 5, 131072 bytes)
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] Inode-cache hash table entries: 16384 (order: 4, 65536 bytes)
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] Memory: 192MB = 192MB total
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] Memory: 188976k/188976k available, 7632k reserved, 0K highmem
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] Virtual kernel memory layout:
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] vector : 0xffff0000 - 0xffff1000 ( 4 kB)
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] fixmap : 0xfff00000 - 0xfffe0000 ( 896 kB)
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] vmalloc : 0xcc800000 - 0xe8000000 ( 440 MB)
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] lowmem : 0xc0000000 - 0xcc000000 ( 192 MB)
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] modules : 0xbf000000 - 0xc0000000 ( 16 MB)
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] .text : 0xc0008000 - 0xc04c0e78 (4836 kB)
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] .init : 0xc04c1000 - 0xc04e0b10 ( 127 kB)
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] .data : 0xc04e2000 - 0xc050e1c0 ( 177 kB)
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] .bss : 0xc050e1e4 - 0xc05b5128 ( 668 kB)
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] NR_IRQS:330
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] sched_clock: 32 bits at 1000kHz, resolution 1000ns, wraps every 4294967ms
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] timer_set_mode: unhandled mode:1
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] timer_set_mode: unhandled mode:3
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] Console: colour dummy device 80x30
Oct 28 22:59:40 raspberrypi kernel: [ 0.000000] console [tty1] enabled
Oct 28 22:59:40 raspberrypi kernel: [ 0.000945] Calibrating delay loop... 697.95 BogoMIPS (lpj=3489792)
Oct 28 22:59:40 raspberrypi kernel: [ 0.060149] pid_max: default: 32768 minimum: 301
Oct 28 22:59:40 raspberrypi kernel: [ 0.060624] Mount-cache hash table entries: 512
Oct 28 22:59:40 raspberrypi kernel: [ 0.061110] Initializing cgroup subsys cpuacct
Oct 28 22:59:40 raspberrypi kernel: [ 0.061201] Initializing cgroup subsys devices
Oct 28 22:59:40 raspberrypi kernel: [ 0.061237] Initializing cgroup subsys freezer
Oct 28 22:59:40 raspberrypi kernel: [ 0.061267] Initializing cgroup subsys blkio
Oct 28 22:59:40 raspberrypi kernel: [ 0.061414] CPU: Testing write buffer coherency: ok
Oct 28 22:59:40 raspberrypi kernel: [ 0.061766] hw perfevents: enabled with v6 PMU driver, 3 counters available
Oct 28 22:59:40 raspberrypi kernel: [ 0.062987] devtmpfs: initialized
Oct 28 22:59:40 raspberrypi kernel: [ 0.073382] NET: Registered protocol family 16
Oct 28 22:59:40 raspberrypi kernel: [ 0.074070] bcm2708.uart_clock = 0
Oct 28 22:59:40 raspberrypi kernel: [ 0.075689] hw-breakpoint: found 6 breakpoint and 1 watchpoint registers.
Oct 28 22:59:40 raspberrypi kernel: [ 0.075745] hw-breakpoint: maximum watchpoint size is 4 bytes.
Oct 28 22:59:40 raspberrypi kernel: [ 0.075783] mailbox: Broadcom VideoCore Mailbox driver
Oct 28 22:59:40 raspberrypi kernel: [ 0.075890] bcm2708_vcio: mailbox at f200b880
Oct 28 22:59:40 raspberrypi kernel: [ 0.075995] bcm_power: Broadcom power driver
Oct 28 22:59:40 raspberrypi kernel: [ 0.076034] bcm_power_open() -> 0
Oct 28 22:59:40 raspberrypi kernel: [ 0.076060] bcm_power_request(0, 8)
Oct 28 22:59:40 raspberrypi kernel: [ 0.576739] bcm_mailbox_read -> 00000080, 0
Oct 28 22:59:40 raspberrypi kernel: [ 0.576780] bcm_power_request -> 0
Oct 28 22:59:40 raspberrypi kernel: [ 0.576805] Serial: AMBA PL011 UART driver
Oct 28 22:59:40 raspberrypi kernel: [ 0.576963] dev:f1: ttyAMA0 at MMIO 0x20201000 (irq = 83) is a PL011 rev3
Oct 28 22:59:40 raspberrypi kernel: [ 0.886637] console [ttyAMA0] enabled
Oct 28 22:59:40 raspberrypi kernel: [ 0.912141] bio: create slab <bio-0> at 0
Oct 28 22:59:40 raspberrypi kernel: [ 0.917368] SCSI subsystem initialized
Oct 28 22:59:40 raspberrypi kernel: [ 0.921363] usbcore: registered new interface driver usbfs
Oct 28 22:59:40 raspberrypi kernel: [ 0.927096] usbcore: registered new interface driver hub
Oct 28 22:59:40 raspberrypi kernel: [ 0.932682] usbcore: registered new device driver usb
Oct 28 22:59:40 raspberrypi kernel: [ 0.938733] Switching to clocksource stc
Oct 28 22:59:40 raspberrypi kernel: [ 0.942982] FS-Cache: Loaded
Oct 28 22:59:40 raspberrypi kernel: [ 0.946164] CacheFiles: Loaded
Oct 28 22:59:40 raspberrypi kernel: [ 0.962875] NET: Registered protocol family 2
Oct 28 22:59:40 raspberrypi kernel: [ 0.967705] IP route cache hash table entries: 2048 (order: 1, 8192 bytes)
Oct 28 22:59:40 raspberrypi kernel: [ 0.975776] TCP established hash table entries: 8192 (order: 4, 65536 bytes)
Oct 28 22:59:40 raspberrypi kernel: [ 0.983275] TCP bind hash table entries: 8192 (order: 3, 32768 bytes)
Oct 28 22:59:40 raspberrypi kernel: [ 0.989918] TCP: Hash tables configured (established 8192 bind 8192)
Oct 28 22:59:40 raspberrypi kernel: [ 0.996337] TCP reno registered
Oct 28 22:59:40 raspberrypi kernel: [ 0.999506] UDP hash table entries: 256 (order: 0, 4096 bytes)
Oct 28 22:59:40 raspberrypi kernel: [ 1.005421] UDP-Lite hash table entries: 256 (order: 0, 4096 bytes)
Oct 28 22:59:40 raspberrypi kernel: [ 1.012208] NET: Registered protocol family 1
Oct 28 22:59:40 raspberrypi kernel: [ 1.017397] RPC: Registered named UNIX socket transport module.
Oct 28 22:59:40 raspberrypi kernel: [ 1.023524] RPC: Registered udp transport module.
Oct 28 22:59:40 raspberrypi kernel: [ 1.028248] RPC: Registered tcp transport module.
Oct 28 22:59:40 raspberrypi kernel: [ 1.033000] RPC: Registered tcp NFSv4.1 backchannel transport module.
Oct 28 22:59:40 raspberrypi kernel: [ 1.039745] bcm2708_dma: DMA manager at cc808000
Oct 28 22:59:40 raspberrypi kernel: [ 1.044546] bcm2708_gpio: bcm2708_gpio_probe c04e9c18
Oct 28 22:59:40 raspberrypi kernel: [ 1.050005] vc-mem: phys_addr:0x00000000 mem_base=0x0c000000 mem_size:0x10000000(256 MiB)
Oct 28 22:59:40 raspberrypi kernel: [ 1.059299] audit: initializing netlink socket (disabled)
Oct 28 22:59:40 raspberrypi kernel: [ 1.064849] type=2000 audit(0.750:1): initialized
Oct 28 22:59:40 raspberrypi kernel: [ 1.180968] VFS: Disk quotas dquot_6.5.2
Oct 28 22:59:40 raspberrypi kernel: [ 1.185212] Dquot-cache hash table entries: 1024 (order 0, 4096 bytes)
Oct 28 22:59:40 raspberrypi kernel: [ 1.192748] FS-Cache: Netfs 'nfs' registered for caching
Oct 28 22:59:40 raspberrypi kernel: [ 1.199017] msgmni has been set to 369
Oct 28 22:59:40 raspberrypi kernel: [ 1.204875] io scheduler noop registered
Oct 28 22:59:40 raspberrypi kernel: [ 1.208844] io scheduler deadline registered (default)
Oct 28 22:59:40 raspberrypi kernel: [ 1.214287] io scheduler cfq registered
Oct 28 22:59:40 raspberrypi kernel: [ 1.233522] Console: switching to colour frame buffer device 82x26
Oct 28 22:59:40 raspberrypi kernel: [ 1.245823] kgdb: Registered I/O driver kgdboc.
Oct 28 22:59:40 raspberrypi kernel: [ 1.262200] brd: module loaded
Oct 28 22:59:40 raspberrypi kernel: [ 1.272520] loop: module loaded
Oct 28 22:59:40 raspberrypi kernel: [ 1.278715] vcos: [1]: vchiq_init_state: slot_zero = 0xffd80000, is_master = 0
Oct 28 22:59:40 raspberrypi kernel: [ 1.288447] vcos: [1]: vchiq_init_state: called
Oct 28 22:59:40 raspberrypi kernel: [ 1.296179] vcos: [1]: vchiq: initialised - version 2 (min 2), device 252.0
Oct 28 22:59:40 raspberrypi kernel: [ 1.306171] usbcore: registered new interface driver smsc95xx
Oct 28 22:59:40 raspberrypi kernel: [ 1.314280] cdc_ncm: 04-Aug-2011
Oct 28 22:59:40 raspberrypi kernel: [ 1.319826] usbcore: registered new interface driver cdc_ncm
Oct 28 22:59:40 raspberrypi kernel: [ 1.327776] dwc_otg: version 3.00a 10-AUG-2012 (platform bus)
Oct 28 22:59:40 raspberrypi kernel: [ 1.540891] Core Release: 2.80a
Oct 28 22:59:40 raspberrypi kernel: [ 1.546306] Setting default values for core params
Oct 28 22:59:40 raspberrypi kernel: [ 1.553309] Finished setting default values for core params
Oct 28 22:59:40 raspberrypi kernel: [ 1.766162] Using Buffer DMA mode
Oct 28 22:59:40 raspberrypi kernel: [ 1.771579] Periodic Transfer Interrupt Enhancement - disabled
Oct 28 22:59:40 raspberrypi kernel: [ 1.779570] Multiprocessor Interrupt Enhancement - disabled
Oct 28 22:59:40 raspberrypi kernel: [ 1.787288] OTG VER PARAM: 0, OTG VER FLAG: 0
Oct 28 22:59:40 raspberrypi kernel: [ 1.793835] Dedicated Tx FIFOs mode
Oct 28 22:59:40 raspberrypi kernel: [ 1.799647] dwc_otg: Microframe scheduler enabled
Oct 28 22:59:40 raspberrypi kernel: [ 1.799701] dwc_otg bcm2708_usb: DWC OTG Controller
Oct 28 22:59:40 raspberrypi kernel: [ 1.806896] dwc_otg bcm2708_usb: new USB bus registered, assigned bus number 1
Oct 28 22:59:40 raspberrypi kernel: [ 1.816540] dwc_otg bcm2708_usb: irq 32, io mem 0x00000000
Oct 28 22:59:40 raspberrypi kernel: [ 1.824297] Init: Port Power? op_state=1
Oct 28 22:59:40 raspberrypi kernel: [ 1.830339] Init: Power Port (0)
Oct 28 22:59:40 raspberrypi kernel: [ 1.835837] usb usb1: New USB device found, idVendor=1d6b, idProduct=0002
Oct 28 22:59:40 raspberrypi kernel: [ 1.844898] usb usb1: New USB device strings: Mfr=3, Product=2, SerialNumber=1
Oct 28 22:59:40 raspberrypi kernel: [ 1.854347] usb usb1: Product: DWC OTG Controller
Oct 28 22:59:40 raspberrypi kernel: [ 1.861211] usb usb1: Manufacturer: Linux 3.2.27+ dwc_otg_hcd
Oct 28 22:59:40 raspberrypi kernel: [ 1.869140] usb usb1: SerialNumber: bcm2708_usb
Oct 28 22:59:40 raspberrypi kernel: [ 1.876712] hub 1-0:1.0: USB hub found
Oct 28 22:59:40 raspberrypi kernel: [ 1.882598] hub 1-0:1.0: 1 port detected
Oct 28 22:59:40 raspberrypi kernel: [ 1.889349] dwc_otg: FIQ enabled
Oct 28 22:59:40 raspberrypi kernel: [ 1.889368] dwc_otg: NAK holdoff enabled
Oct 28 22:59:40 raspberrypi kernel: [ 1.889389] Module dwc_common_port init
Oct 28 22:59:40 raspberrypi kernel: [ 1.889731] usbcore: registered new interface driver uas
Oct 28 22:59:40 raspberrypi kernel: [ 1.897232] Initializing USB Mass Storage driver...
Oct 28 22:59:40 raspberrypi kernel: [ 1.904460] usbcore: registered new interface driver usb-storage
Oct 28 22:59:40 raspberrypi kernel: [ 1.912540] USB Mass Storage support registered.
Oct 28 22:59:40 raspberrypi kernel: [ 1.919488] usbcore: registered new interface driver libusual
Oct 28 22:59:40 raspberrypi kernel: [ 1.927783] mousedev: PS/2 mouse device common for all mice
Oct 28 22:59:40 raspberrypi kernel: [ 1.936377] bcm2835-cpufreq: min=700000 max=700000 cur=700000
Oct 28 22:59:40 raspberrypi kernel: [ 1.942234] bcm2835-cpufreq: switching to governor powersave
Oct 28 22:59:40 raspberrypi kernel: [ 1.950044] bcm2835-cpufreq: switching to governor powersave
Oct 28 22:59:40 raspberrypi kernel: [ 1.957919] cpuidle: using governor ladder
Oct 28 22:59:40 raspberrypi kernel: [ 1.966057] cpuidle: using governor menu
Oct 28 22:59:40 raspberrypi kernel: [ 1.971989] sdhci: Secure Digital Host Controller Interface driver
Oct 28 22:59:40 raspberrypi kernel: [ 1.980182] sdhci: Copyright(c) Pierre Ossman
Oct 28 22:59:40 raspberrypi kernel: [ 1.986789] sdhci: Enable low-latency mode
Oct 28 22:59:40 raspberrypi kernel: [ 1.992928] bcm_power_open() -> 1
Oct 28 22:59:40 raspberrypi kernel: [ 1.998385] Registered led device: mmc0::
Oct 28 22:59:40 raspberrypi kernel: [ 1.999856] mmc0: SDHCI controller on BCM2708_Arasan [platform] using platform's DMA
Oct 28 22:59:40 raspberrypi kernel: [ 2.011694] mmc0: BCM2708 SDHC host at 0x20300000 DMA 2 IRQ 77
Oct 28 22:59:40 raspberrypi kernel: [ 2.019845] sdhci-pltfm: SDHCI platform and OF driver helper
Oct 28 22:59:40 raspberrypi kernel: [ 2.029045] Registered led device: led0
Oct 28 22:59:40 raspberrypi kernel: [ 2.029444] usbcore: registered new interface driver usbhid
Oct 28 22:59:40 raspberrypi kernel: [ 2.037204] usbhid: USB HID core driver
Oct 28 22:59:40 raspberrypi kernel: [ 2.054110] TCP cubic registered
Oct 28 22:59:40 raspberrypi kernel: [ 2.061588] Initializing XFRM netlink socket
Oct 28 22:59:40 raspberrypi kernel: [ 2.070167] NET: Registered protocol family 17
Oct 28 22:59:40 raspberrypi kernel: [ 2.078947] Registering the dns_resolver key type
Oct 28 22:59:40 raspberrypi kernel: [ 2.087971] Indeed it is in host mode hprt0 = 00021501
Oct 28 22:59:40 raspberrypi kernel: [ 2.095361] VFP support v0.3: implementor 41 architecture 1 part 20 variant b rev 5
Oct 28 22:59:40 raspberrypi kernel: [ 2.123632] registered taskstats version 1
Oct 28 22:59:40 raspberrypi kernel: [ 2.130011] drivers/rtc/hctosys.c: unable to open rtc device (rtc0)
Oct 28 22:59:40 raspberrypi kernel: [ 2.148071] mmc0: new high speed SD card at address 0002
Oct 28 22:59:40 raspberrypi kernel: [ 2.163236] Waiting for root device /dev/mmcblk0p2...
Oct 28 22:59:40 raspberrypi kernel: [ 2.171375] mmcblk0: mmc0:0002 N/A 1.80 GiB
Oct 28 22:59:40 raspberrypi kernel: [ 2.194145] mmcblk0: p1 p2
Oct 28 22:59:40 raspberrypi kernel: [ 2.286609] EXT4-fs (mmcblk0p2): mounted filesystem with ordered data mode. Opts: (null)
Oct 28 22:59:40 raspberrypi kernel: [ 2.299371] VFS: Mounted root (ext4 filesystem) on device 179:2.
Oct 28 22:59:40 raspberrypi kernel: [ 2.316849] devtmpfs: mounted
Oct 28 22:59:40 raspberrypi kernel: [ 2.323047] Freeing init memory: 124K
Oct 28 22:59:40 raspberrypi kernel: [ 2.332825] usb 1-1: new high-speed USB device number 2 using dwc_otg
Oct 28 22:59:40 raspberrypi kernel: [ 2.341754] Indeed it is in host mode hprt0 = 00001101
Oct 28 22:59:40 raspberrypi kernel: [ 2.553127] usb 1-1: New USB device found, idVendor=0424, idProduct=9512
Oct 28 22:59:40 raspberrypi kernel: [ 2.562659] usb 1-1: New USB device strings: Mfr=0, Product=0, SerialNumber=0
Oct 28 22:59:40 raspberrypi kernel: [ 2.573229] hub 1-1:1.0: USB hub found
Oct 28 22:59:40 raspberrypi kernel: [ 2.579794] hub 1-1:1.0: 3 ports detected
Oct 28 22:59:40 raspberrypi kernel: [ 2.862976] usb 1-1.1: new high-speed USB device number 3 using dwc_otg
Oct 28 22:59:40 raspberrypi kernel: [ 2.993359] usb 1-1.1: New USB device found, idVendor=0424, idProduct=ec00
Oct 28 22:59:40 raspberrypi kernel: [ 3.003025] usb 1-1.1: New USB device strings: Mfr=0, Product=0, SerialNumber=0
Oct 28 22:59:40 raspberrypi kernel: [ 3.020538] smsc95xx v1.0.4
Oct 28 22:59:40 raspberrypi kernel: [ 3.098302] smsc95xx 1-1.1:1.0: eth0: register 'smsc95xx' at usb-bcm2708_usb-1.1, smsc95xx USB 2.0 Ethernet, b8:27:eb:a9:38:eb
Oct 28 22:59:40 raspberrypi kernel: [ 12.098183] EXT4-fs (mmcblk0p2): re-mounted. Opts: (null)
Oct 28 22:59:40 raspberrypi kernel: [ 12.664302] EXT4-fs (mmcblk0p2): re-mounted. Opts: (null)
Oct 28 22:59:40 raspberrypi kernel: [ 13.572103] bcm2835 ALSA card created!
Oct 28 22:59:40 raspberrypi kernel: [ 13.586854] bcm2835 ALSA chip created!
Oct 28 22:59:40 raspberrypi kernel: [ 13.598656] bcm2835 ALSA chip created!
Oct 28 22:59:40 raspberrypi kernel: [ 13.608965] bcm2835 ALSA chip created!
Oct 28 22:59:40 raspberrypi kernel: [ 13.617221] bcm2835 ALSA chip created!
Oct 28 22:59:40 raspberrypi kernel: [ 13.625431] bcm2835 ALSA chip created!
Oct 28 22:59:40 raspberrypi kernel: [ 13.633524] bcm2835 ALSA chip created!
Oct 28 22:59:40 raspberrypi kernel: [ 13.641442] bcm2835 ALSA chip created!
Oct 28 22:59:40 raspberrypi kernel: [ 21.881724] smsc95xx 1-1.1:1.0: eth0: link up, 100Mbps, full-duplex, lpa 0x45E1
Oct 28 22:59:44 raspberrypi kernel: [ 30.121038] Adding 102396k swap on /var/swap. Priority:-1 extents:1 across:102396k SS


This is my cmdline.txt:
dwc_otg.lpm_enable=0 smsc95xx.turbo_mode=N console=ttyAMA0,115200 kgdboc=ttyAMA0,115200 console=tty1 root=/dev/mmcblk0p2 rootfstype=ext4 elevator=deadline rootwait


Fails both with smsc95xx.turbo_mode=N and without it...

64 bytes from raspberry (172.16.1.44): icmp_req=278 ttl=64 time=1.94 ms
64 bytes from raspberry (172.16.1.44): icmp_req=279 ttl=64 time=1.63 ms
64 bytes from raspberry (172.16.1.44): icmp_req=280 ttl=64 time=0.938 ms
64 bytes from raspberry (172.16.1.44): icmp_req=281 ttl=64 time=1.70 ms
64 bytes from raspberry (172.16.1.44): icmp_req=282 ttl=64 time=1.00 ms
64 bytes from raspberry (172.16.1.44): icmp_req=283 ttl=64 time=1.13 ms
64 bytes from raspberry (172.16.1.44): icmp_req=284 ttl=64 time=1.73 ms
64 bytes from raspberry (172.16.1.44): icmp_req=285 ttl=64 time=5.96 ms
64 bytes from raspberry (172.16.1.44): icmp_req=286 ttl=64 time=3.88 ms
64 bytes from raspberry (172.16.1.44): icmp_req=287 ttl=64 time=1.82 ms
64 bytes from raspberry (172.16.1.44): icmp_req=288 ttl=64 time=72.4 ms
64 bytes from raspberry (172.16.1.44): icmp_req=289 ttl=64 time=1.55 ms
64 bytes from raspberry (172.16.1.44): icmp_req=290 ttl=64 time=1.71 ms
64 bytes from raspberry (172.16.1.44): icmp_req=291 ttl=64 time=1.53 ms
64 bytes from raspberry (172.16.1.44): icmp_req=292 ttl=64 time=1.98 ms
64 bytes from raspberry (172.16.1.44): icmp_req=293 ttl=64 time=1.28 ms
64 bytes from raspberry (172.16.1.44): icmp_req=294 ttl=64 time=2.90 ms
64 bytes from raspberry (172.16.1.44): icmp_req=295 ttl=64 time=1.42 ms
64 bytes from raspberry (172.16.1.44): icmp_req=296 ttl=64 time=1.89 ms
64 bytes from raspberry (172.16.1.44): icmp_req=297 ttl=64 time=1.88 ms
64 bytes from raspberry (172.16.1.44): icmp_req=298 ttl=64 time=1.76 ms
64 bytes from raspberry (172.16.1.44): icmp_req=299 ttl=64 time=0.938 ms
64 bytes from raspberry (172.16.1.44): icmp_req=300 ttl=64 time=1.39 ms
64 bytes from raspberry (172.16.1.44): icmp_req=301 ttl=64 time=1.20 ms
64 bytes from raspberry (172.16.1.44): icmp_req=302 ttl=64 time=1.54 ms
64 bytes from raspberry (172.16.1.44): icmp_req=303 ttl=64 time=1.60 ms
64 bytes from raspberry (172.16.1.44): icmp_req=304 ttl=64 time=1.85 ms
64 bytes from raspberry (172.16.1.44): icmp_req=305 ttl=64 time=1.31 ms
From getsugawifi (172.16.1.133) icmp_seq=343 Destination Host Unreachable
From getsugawifi (172.16.1.133) icmp_seq=344 Destination Host Unreachable
From getsugawifi (172.16.1.133) icmp_seq=347 Destination Host Unreachable
From getsugawifi (172.16.1.133) icmp_seq=348 Destination Host Unreachable
From getsugawifi (172.16.1.133) icmp_seq=349 Destination Host Unreachable
From getsugawifi (172.16.1.133) icmp_seq=350 Destination Host Unreachable
From getsugawifi (172.16.1.133) icmp_seq=351 Destination Host Unreachable
From getsugawifi (172.16.1.133) icmp_seq=352 Destination Host Unreachable
From getsugawifi (172.16.1.133) icmp_seq=353 Destination Host Unreachable
From getsugawifi (172.16.1.133) icmp_seq=354 Destination Host Unreachable
From getsugawifi (172.16.1.133) icmp_seq=355 Destination Host Unreachable
From getsugawifi (172.16.1.133) icmp_seq=356 Destination Host Unreachable
From getsugawifi (172.16.1.133) icmp_seq=357 Destination Host Unreachable
From getsugawifi (172.16.1.133) icmp_seq=358 Destination Host Unreachable
From getsugawifi (172.16.1.133) icmp_seq=359 Destination Host Unreachable
From getsugawifi (172.16.1.133) icmp_seq=360 Destination Host Unreachable
From getsugawifi (172.16.1.133) icmp_seq=361 Destination Host Unreachable
From getsugawifi (172.16.1.133) icmp_seq=362 Destination Host Unreachable
From getsugawifi (172.16.1.133) icmp_seq=364 Destination Host Unreachable
From getsugawifi (172.16.1.133) icmp_seq=365 Destination Host Unreachable
From getsugawifi (172.16.1.133) icmp_seq=366 Destination Host Unreachable
From getsugawifi (172.16.1.133) icmp_seq=367 Destination Host Unreachable
^C
--- raspberry ping statistics ---
368 packets transmitted, 305 received, +22 errors, 17% packet loss, time 367474ms
rtt min/avg/max/mdev = 0.806/2.582/119.058/10.703 ms, pipe 4


Measured 4.93v between TP1 and TP2 when it hanged.

Should I RMA it? :-(
User avatar
Posts: 6
Joined: Thu Nov 01, 2012 8:54 pm
by gsh » Sat Nov 03, 2012 9:45 pm
--
Gordon Hollingworth PhD
Raspberry Pi - Director of Software Engineering
Raspberry Pi Engineer & Forum Moderator
Raspberry Pi Engineer & Forum Moderator
Posts: 789
Joined: Sat Sep 10, 2011 11:43 am
by jesusdf » Tue Nov 06, 2012 8:32 am
Thank you for your response, I will try it this evening and give feedback! :)
User avatar
Posts: 6
Joined: Thu Nov 01, 2012 8:54 pm
by arsene » Tue Nov 06, 2012 5:31 pm
Hello everyone,

Using Rpi Model B with 2012-10-28-wheezy-raspbian.img . Please forgive my poor English.

I had the same problems. Since the power supply seemed to be the origin of the problems, according to this topic, I decided to change mine.

So I left a basic trasformer which outputs 5V 1A and used instead a PowerBank (TeckNet iEP 392) to supply the Rpi with stabilized current (and which was aimed at being used with a PirateBox).

The PowerBank is capable of 5V 2.1A and won't let any drop in voltage, even when heavily using my NIC with USB mouse and keyboard connected.

Since then, no more freezes or hangs! I can finally
Code: Select all
sudo apt-get update
and
Code: Select all
sudo apt-get dist-upgrade
successfully (which were my first concerns)!

Hope this helps.
Posts: 2
Joined: Mon Nov 05, 2012 8:48 pm
by gsh » Tue Nov 06, 2012 8:15 pm
ggodart wrote:Not sure if this is related, but I'm setting up my PI and installing stuff with apt-get, this fails when connecting to the mirrordirector, after which the network goes down and I cannot even ping anything, network restart has no effect and I have to reboot.
I can workaround this by running a ping session in another window in parallel to and before the apt-get, I suspect this is a software issue and probably a timeout. I'm in the UK and connects to the mirrordirector are really variable - ping takes between 50ms and 3,000ms.


We have a fix for a problem like this, you should try updating and retrying
--
Gordon Hollingworth PhD
Raspberry Pi - Director of Software Engineering
Raspberry Pi Engineer & Forum Moderator
Raspberry Pi Engineer & Forum Moderator
Posts: 789
Joined: Sat Sep 10, 2011 11:43 am
by arsene » Thu Nov 15, 2012 2:01 am
gsh wrote:
ggodart wrote:Not sure if this is related, but I'm setting up my PI and installing stuff with apt-get, this fails when connecting to the mirrordirector, after which the network goes down and I cannot even ping anything, network restart has no effect and I have to reboot.
I can workaround this by running a ping session in another window in parallel to and before the apt-get, I suspect this is a software issue and probably a timeout. I'm in the UK and connects to the mirrordirector are really variable - ping takes between 50ms and 3,000ms.


We have a fix for a problem like this, you should try updating and retrying


Same problem as well. I cannot tell which from the down server or my underfed Pi would cause the lost of connection. But when I tried to get to the Repositories with my laptop's browser, it timedout as well.

Because I did not want to wait for the server to be repared or the connection to be available again, I changed the source and nowadays use the ones supplied by the Raspian community. Check out the Repository and Mirror links on http://www.raspbian.org/
Posts: 2
Joined: Mon Nov 05, 2012 8:48 pm
by da_maddin » Fri Dec 14, 2012 1:24 am
Hi,
I experienced the same network outage issue - PI was stable but Network LED's would go off and no response from the keyboard, getting worse and worse the longer the thing ran.

I went through every PSU Option I have (including power via GPIO) to no avail and you brought me to the right conclusion. It actually IS heat in my case. The network chip heats up and that drops the line. If i place my finger on it its giving that "little sting" of heat exchange and the LED's light up again.

I placed a fan over it and it has not dropped any packet ever since. Before it was unusable.

Hope this helps.
maddin
PS: Its probably the chip not being properly soldered but i am not sending it in, i need it right now so a fan will be installed.
Posts: 1
Joined: Fri Dec 14, 2012 12:57 am
by durian » Fri Dec 14, 2012 10:40 am
I had a problem with wifi (tplink USB adapter, 8192cu module) which would drop wifi under high traffic (rsync, pacman -Syu), on a Pi model B running Arch. Wifi would disappear, but the Pi continued working, even the USB webcam. I added a powered HUB and tested without the webcam, to no avail.

Setting mtu to 1000 instead of 1500 on wlan0 appears to have solved the problem (knock on wood &c). I haven't had problems since (2 days ago).

-peter
Posts: 1
Joined: Thu Dec 13, 2012 6:38 pm
by jives11 » Wed Dec 26, 2012 9:32 am
Interesting thread. Am seeing simlar problems. I decided to use my Pi as a headless rsync server over ethernet. I attached an external , powered 1Tb Hitachi USB drive, which I formatted and mounted in /etc/fstab. I then installed the rsync server and configured rsyncd.conf


When I kick off an rsync from a remote system, its runs fine for about 10 minutes, then freezes. the Pi doesn't respond to ssh or ping. If I reboot the pi and restart the rsync, same thing. copies more files across then freezes after roughly 10 minutes.

So I have read this thread. Is the consensus that returning the Pi for another unit might fix the issue OR should I try and use a different USB drive. The one I'm using is powered so presumably drawing little or no current via the USB interface ?
Posts: 6
Joined: Fri Jul 20, 2012 7:49 am
by NewPi » Tue Jan 01, 2013 2:43 pm
durian wrote:I had a problem with wifi (tplink USB adapter, 8192cu module) which would drop wifi under high traffic (rsync, pacman -Syu), on a Pi model B running Arch. Wifi would disappear, but the Pi continued working, even the USB webcam. I added a powered HUB and tested without the webcam, to no avail.

Setting mtu to 1000 instead of 1500 on wlan0 appears to have solved the problem (knock on wood &c). I haven't had problems since (2 days ago).

-peter


Can you pls. tell how to change MTU, I am using "Network Manager" applet.
Raspberry Pi Howto, Tips, Tricks and Tools -> http://bit.ly/RPiTricks
Posts: 66
Joined: Sat Aug 18, 2012 2:52 pm
by jives11 » Fri Jan 04, 2013 6:17 pm
Well farnell shipped me a new Pi on an RMA basis over night (very good service), and this is working fine. Now USB/network issues and my pi is working nicely as an rsync server.

What was interesting was that I needed to re burn the SD card. The old OS image I was using with my old pi didn't work with the new one , even though it had been apt-get upgraded a few days earlier. When powered on I got red light, no green and no amber/orange lights. When I re imaged the SD card with the latest OS download it started fine. Its a lot easier now with the raspi-config tool as I could extend my root partition easily to fill the SD card. I could also disable a GUI and sshd seemed to be enabled by default, so much easier to provision headless pi's
Posts: 6
Joined: Fri Jul 20, 2012 7:49 am
by eshatohhin » Thu Jan 10, 2013 7:54 pm
I'm having related problem.
I have two 2tb drives plugged into raspberry. Both drives are self powered and attached directly to raspberry.
Under heavy concurrent disk load I get this in log every few minutes:
Code: Select all
usb 1-1.3: reset high-speed USB device number 4 using dwc_otg

and then after about 10 of these in a row system hangs completely.

If i read from two drives - no problems, tested with
Code: Select all
cat /dev/sda1 > /dev/null &
cat /dev/sdb1 > /dev/null &


But when i try to write to two drives - problem occurs:
Code: Select all
dd if=/dev/zero of=/mnt/disk1/test bs=1M count=10000 &
dd if=/dev/zero of=/mnt/disk2/test bs=1M count=10000 &


Sometimes i also get this in log too:
Code: Select all
smsc95xx 1-1.1:1.0: eth0: Failed to read register index 0x00000118


have tried different power supplies (2A), cables, etc. - same result
tried different drives - same result

played with buffer size with no success
Code: Select all
echo 1 > /proc/sys/vm/dirty_background_ratio

Code: Select all
echo 20 > /proc/sys/vm/dirty_background_ratio


What i've observed from "atop", before i get this message in log, drive sits in 102%+ busy for several seconds,
then this reset happens

First i've found this while playing with software raid on two drives, i was getting almost 100% stall rate when it tries to sync for the first time:
Code: Select all
dd if=/dev/zero of=/mnt/d1/raid bs=1M count=1000
dd if=/dev/zero of=/mnt/d2/raid bs=1M count=1000
losetup /dev/loop0 /mnt/d1/raid
losetup /dev/loop1 /mnt/d2/raid
mdadm --create --verbose /dev/md0 --level=raid1 --raid-devices=2 /dev/loop0 /dev/loop1
User avatar
Posts: 11
Joined: Tue May 15, 2012 12:27 pm
Location: Estonia
by rpifreeze » Tue Jan 15, 2013 6:08 pm
jives11 wrote: Now USB/network issues and my pi is working nicely as an rsync server.

Come on guys, be exact. Rsync is not meant to measure speed which might be 100kB/s or 10 MB/s. It depends on many factors, like amount of files on one folder, average file size and so on.
What speed did you get ?
Also, were there writes to sdcard or disk. Or maybe reads ?
I faced the same problem with "cannot register indexes..." when sd card was either damaged (filesystem) or when tens of thousands of file resided in one folder. In that case, during read, system crashed/freezed. The same occured on HDD. I tried to change filesystem to ext3 and even trying to disable journaling. Not worse, but little better - ext2 works the best and I get very rare crashes.
usb and broadcom issues will never be solved...Too much time passed, too much to be done...
Posts: 63
Joined: Tue Jul 24, 2012 10:03 pm