EnfermeraSexy
Posts: 17
Joined: Mon Jan 02, 2017 6:19 pm

Stuttering when writing to a USB

Mon Jan 02, 2017 7:04 pm

Hello! I bought a Raspberry Pi 3 a couple days ago. I want to have it as a downloader manager, but i have a problem. I'll try to describe it.
For example, i've got a folder in a USB that i share with samba. If i copy/write something to it it seems to stutter, for example you see that it is writing with a speed of 8 MB/s and every 5 seconds (it doesn't have to be 5, it can be 10, 3... you know without regular intervals) it drops to 0 for over 1-5 seconds and backs again 8.
It's not just samba, happens with Transmission too, if i download something i see how it stucks and keeps downloading after some seconds, it happens with every program. It's not the USB (it happens too if i use a folder in SD), on Windows i get like 130 MB/s reading and 50 writing (i know Raspberry only has usb 2.0). I tried another USB (diferent) and nothing changes.

I tried a lot of filesystems (NTFS, FAT32, EXT4, EXT3, EXT2, ExFat) and seems like FAT32 it's the best, doesn't drops speed but you know... i can use every format except it (4 GB limits). The worst seems to be EXT2, all EXT works really bad, because NTFS works well, it drops sometimes, but not so much as EXT.
If you are on Raspberry it stucks too, i mean i had a open document, and when it got stutters i tried to save it, but it got stucks when i clicked on save, and after a few seconds it continue writing and saved the document.
CPU usage isn't high, about 10-15%, maybe 5% more if i use NTFS.
I'm using Raspbian with latest updates and firmware. I tried making it boot from USB but it stills happening, less but happens.

Sorry for my english, i know i have to improve it >.<

ejolson
Posts: 5137
Joined: Tue Mar 18, 2014 11:47 am

Re: Stuttering when writing to a USB

Mon Jan 02, 2017 10:58 pm

EnfermeraSexy wrote:Hello! I bought a Raspberry Pi 3 a couple days ago. I want to have it as a downloader manager, but i have a problem. I'll try to describe it.
For example, i've got a folder in a USB that i share with samba. If i copy/write something to it it seems to stutter, for example you see that it is writing with a speed of 8 MB/s and every 5 seconds (it doesn't have to be 5, it can be 10, 3... you know without regular intervals) it drops to 0 for over 1-5 seconds and backs again 8.
It's not just samba, happens with Transmission too, if i download something i see how it stucks and keeps downloading after some seconds, it happens with every program. It's not the USB (it happens too if i use a folder in SD), on Windows i get like 130 MB/s reading and 50 writing (i know Raspberry only has usb 2.0). I tried another USB (diferent) and nothing changes.

I tried a lot of filesystems (NTFS, FAT32, EXT4, EXT3, EXT2, ExFat) and seems like FAT32 it's the best, doesn't drops speed but you know... i can use every format except it (4 GB limits). The worst seems to be EXT2, all EXT works really bad, because NTFS works well, it drops sometimes, but not so much as EXT.
If you are on Raspberry it stucks too, i mean i had a open document, and when it got stutters i tried to save it, but it got stucks when i clicked on save, and after a few seconds it continue writing and saved the document.
CPU usage isn't high, about 10-15%, maybe 5% more if i use NTFS.
I'm using Raspbian with latest updates and firmware. I tried making it boot from USB but it stills happening, less but happens.

Sorry for my english, i know i have to improve it >.<
This sounds like expected behavior related to memory pressure of the full versus empty filesystem buffer cache. In my observation such stuttering was first noticeable in 2.x kernels and gets more pronounced with additional RAM.

On the other hand you may have bad hardware or device drivers with error conditions that time out and then reset. Please check your log files using dmesg to make sure this isn't happening.

EnfermeraSexy
Posts: 17
Joined: Mon Jan 02, 2017 6:19 pm

Re: Stuttering when writing to a USB

Tue Jan 03, 2017 12:29 am

ejolson wrote:This sounds like expected behavior related to memory pressure of the full versus empty filesystem buffer cache. In my observation such stuttering was first noticeable in 2.x kernels and gets more pronounced with additional RAM.

On the other hand you may have bad hardware or device drivers with error conditions that time out and then reset. Please check your log files using dmesg to make sure this isn't happening.
What i'm looking for? I don't really know how does Linux work, i'll attach the log. I was downloading something like for 4 hours, if it helps.

Code: Select all

pi@raspberrypi:~ $ dmesg
[    0.000000] Booting Linux on physical CPU 0x0
[    0.000000] Initializing cgroup subsys cpuset
[    0.000000] Initializing cgroup subsys cpu
[    0.000000] Initializing cgroup subsys cpuacct
[    0.000000] Linux version 4.4.38-v7+ (dc4@dc4-XPS13-9333) (gcc version 4.9.3 (crosstool-NG crosstool-ng-1.22.0-88-g8460611) ) #938 SMP Thu Dec 15 15:22:21 GMT 2016
[    0.000000] CPU: ARMv7 Processor [410fd034] revision 4 (ARMv7), cr=10c5383d
[    0.000000] CPU: PIPT / VIPT nonaliasing data cache, VIPT aliasing instruction cache
[    0.000000] Machine model: Raspberry Pi 3 Model B Rev 1.2
[    0.000000] cma: Reserved 8 MiB at 0x3a800000
[    0.000000] Memory policy: Data cache writealloc
[    0.000000] On node 0 totalpages: 241664
[    0.000000] free_area_init_node: node 0, pgdat 808c4f40, node_mem_map b9fa6000
[    0.000000]   Normal zone: 2124 pages used for memmap
[    0.000000]   Normal zone: 0 pages reserved
[    0.000000]   Normal zone: 241664 pages, LIFO batch:31
[    0.000000] [bcm2709_smp_init_cpus] enter (9520->f3003010)
[    0.000000] [bcm2709_smp_init_cpus] ncores=4
[    0.000000] PERCPU: Embedded 13 pages/cpu @b9f62000 s22592 r8192 d22464 u53248
[    0.000000] pcpu-alloc: s22592 r8192 d22464 u53248 alloc=13*4096
[    0.000000] pcpu-alloc: [0] 0 [0] 1 [0] 2 [0] 3 
[    0.000000] Built 1 zonelists in Zone order, mobility grouping on.  Total pages: 239540
[    0.000000] Kernel command line: 8250.nr_uarts=0 dma.dmachans=0x7f35 bcm2708_fb.fbwidth=656 bcm2708_fb.fbheight=416 bcm2709.boardrev=0xa02082 bcm2709.serial=0x80cd3a54 smsc95xx.macaddr=B8:27:EB:CD:3A:54 bcm2708_fb.fbswap=1 bcm2709.uart_clock=48000000 vc_mem.mem_base=0x3dc00000 vc_mem.mem_size=0x3f000000  dwc_otg.lpm_enable=0 console=ttyS0,115200 console=tty1 root=/dev/mmcblk0p2 rootfstype=ext4 elevator=deadline fsck.repair=yes rootwait
[    0.000000] PID hash table entries: 4096 (order: 2, 16384 bytes)
[    0.000000] Dentry cache hash table entries: 131072 (order: 7, 524288 bytes)
[    0.000000] Inode-cache hash table entries: 65536 (order: 6, 262144 bytes)
[    0.000000] Memory: 939064K/966656K available (6357K kernel code, 432K rwdata, 1716K rodata, 476K init, 764K bss, 19400K reserved, 8192K cma-reserved)
[    0.000000] Virtual kernel memory layout:
    vector  : 0xffff0000 - 0xffff1000   (   4 kB)
    fixmap  : 0xffc00000 - 0xfff00000   (3072 kB)
    vmalloc : 0xbb800000 - 0xff800000   (1088 MB)
    lowmem  : 0x80000000 - 0xbb000000   ( 944 MB)
    modules : 0x7f000000 - 0x80000000   (  16 MB)
      .text : 0x80008000 - 0x807ea688   (8074 kB)
      .init : 0x807eb000 - 0x80862000   ( 476 kB)
      .data : 0x80862000 - 0x808ce250   ( 433 kB)
       .bss : 0x808d1000 - 0x809901ec   ( 765 kB)
[    0.000000] SLUB: HWalign=64, Order=0-3, MinObjects=0, CPUs=4, Nodes=1
[    0.000000] Hierarchical RCU implementation.
[    0.000000] 	Build-time adjustment of leaf fanout to 32.
[    0.000000] NR_IRQS:16 nr_irqs:16 16
[    0.000000] Architected cp15 timer(s) running at 19.20MHz (phys).
[    0.000000] clocksource: arch_sys_counter: mask: 0xffffffffffffff max_cycles: 0x46d987e47, max_idle_ns: 440795202767 ns
[    0.000009] sched_clock: 56 bits at 19MHz, resolution 52ns, wraps every 4398046511078ns
[    0.000027] Switching to timer-based delay loop, resolution 52ns
[    0.000293] Console: colour dummy device 80x30
[    0.001349] console [tty1] enabled
[    0.001395] Calibrating delay loop (skipped), value calculated using timer frequency.. 38.40 BogoMIPS (lpj=192000)
[    0.001462] pid_max: default: 32768 minimum: 301
[    0.001794] Mount-cache hash table entries: 2048 (order: 1, 8192 bytes)
[    0.001838] Mountpoint-cache hash table entries: 2048 (order: 1, 8192 bytes)
[    0.002793] Disabling cpuset control group subsystem
[    0.002850] Initializing cgroup subsys io
[    0.002902] Initializing cgroup subsys memory
[    0.002969] Initializing cgroup subsys devices
[    0.003012] Initializing cgroup subsys freezer
[    0.003055] Initializing cgroup subsys net_cls
[    0.003126] CPU: Testing write buffer coherency: ok
[    0.003214] ftrace: allocating 21237 entries in 63 pages
[    0.051471] CPU0: update cpu_capacity 1024
[    0.051538] CPU0: thread -1, cpu 0, socket 0, mpidr 80000000
[    0.051571] [bcm2709_smp_prepare_cpus] enter
[    0.051726] Setting up static identity map for 0x8240 - 0x8274
[    0.053390] [bcm2709_boot_secondary] cpu:1 started (0) 17
[    0.053582] [bcm2709_secondary_init] enter cpu:1
[    0.053624] CPU1: update cpu_capacity 1024
[    0.053631] CPU1: thread -1, cpu 1, socket 0, mpidr 80000001
[    0.054013] [bcm2709_boot_secondary] cpu:2 started (0) 16
[    0.054177] [bcm2709_secondary_init] enter cpu:2
[    0.054198] CPU2: update cpu_capacity 1024
[    0.054204] CPU2: thread -1, cpu 2, socket 0, mpidr 80000002
[    0.054568] [bcm2709_boot_secondary] cpu:3 started (0) 18
[    0.054699] [bcm2709_secondary_init] enter cpu:3
[    0.054719] CPU3: update cpu_capacity 1024
[    0.054725] CPU3: thread -1, cpu 3, socket 0, mpidr 80000003
[    0.054786] Brought up 4 CPUs
[    0.054885] SMP: Total of 4 processors activated (153.60 BogoMIPS).
[    0.054914] CPU: All CPU(s) started in HYP mode.
[    0.054941] CPU: Virtualization extensions available.
[    0.055580] devtmpfs: initialized
[    0.066498] VFP support v0.3: implementor 41 architecture 3 part 40 variant 3 rev 4
[    0.066873] clocksource: jiffies: mask: 0xffffffff max_cycles: 0xffffffff, max_idle_ns: 19112604462750000 ns
[    0.067610] pinctrl core: initialized pinctrl subsystem
[    0.068155] NET: Registered protocol family 16
[    0.073332] DMA: preallocated 4096 KiB pool for atomic coherent allocations
[    0.080230] hw-breakpoint: found 5 (+1 reserved) breakpoint and 4 watchpoint registers.
[    0.080280] hw-breakpoint: maximum watchpoint size is 8 bytes.
[    0.080454] Serial: AMBA PL011 UART driver
[    0.080606] uart-pl011 3f201000.uart: could not find pctldev for node /soc/gpio@7e200000/uart0_pins, deferring probe
[    0.080807] bcm2835-mbox 3f00b880.mailbox: mailbox enabled
[    0.143613] bcm2835-dma 3f007000.dma: DMA legacy API manager at f3007000, dmachans=0x1
[    0.144218] SCSI subsystem initialized
[    0.144429] usbcore: registered new interface driver usbfs
[    0.144541] usbcore: registered new interface driver hub
[    0.144660] usbcore: registered new device driver usb
[    0.151018] raspberrypi-firmware soc:firmware: Attached to firmware from 2016-12-09 15:11
[    0.178175] clocksource: Switched to clocksource arch_sys_counter
[    0.223209] FS-Cache: Loaded
[    0.223504] CacheFiles: Loaded
[    0.235877] NET: Registered protocol family 2
[    0.236757] TCP established hash table entries: 8192 (order: 3, 32768 bytes)
[    0.236895] TCP bind hash table entries: 8192 (order: 4, 65536 bytes)
[    0.237106] TCP: Hash tables configured (established 8192 bind 8192)
[    0.237221] UDP hash table entries: 512 (order: 2, 16384 bytes)
[    0.237289] UDP-Lite hash table entries: 512 (order: 2, 16384 bytes)
[    0.237539] NET: Registered protocol family 1
[    0.237880] RPC: Registered named UNIX socket transport module.
[    0.237912] RPC: Registered udp transport module.
[    0.237940] RPC: Registered tcp transport module.
[    0.237968] RPC: Registered tcp NFSv4.1 backchannel transport module.
[    0.239066] hw perfevents: enabled with armv7_cortex_a7 PMU driver, 7 counters available
[    0.240422] futex hash table entries: 1024 (order: 4, 65536 bytes)
[    0.253685] VFS: Disk quotas dquot_6.6.0
[    0.254017] VFS: Dquot-cache hash table entries: 1024 (order 0, 4096 bytes)
[    0.256260] FS-Cache: Netfs 'nfs' registered for caching
[    0.257185] NFS: Registering the id_resolver key type
[    0.257247] Key type id_resolver registered
[    0.257276] Key type id_legacy registered
[    0.259622] Block layer SCSI generic (bsg) driver version 0.4 loaded (major 252)
[    0.259777] io scheduler noop registered
[    0.259814] io scheduler deadline registered (default)
[    0.259886] io scheduler cfq registered
[    0.262427] BCM2708FB: allocated DMA memory fac10000
[    0.262474] BCM2708FB: allocated DMA channel 0 @ f3007000
[    0.271201] Console: switching to colour frame buffer device 82x26
[    0.279319] bcm2835-rng 3f104000.rng: hwrng registered
[    0.281719] vc-cma: Videocore CMA driver
[    0.284083] vc-cma: vc_cma_base      = 0x00000000
[    0.286451] vc-cma: vc_cma_size      = 0x00000000 (0 MiB)
[    0.288762] vc-cma: vc_cma_initial   = 0x00000000 (0 MiB)
[    0.291138] vc-mem: phys_addr:0x00000000 mem_base=0x3dc00000 mem_size:0x3f000000(1008 MiB)
[    0.310205] brd: module loaded
[    0.320844] loop: module loaded
[    0.323749] vchiq: vchiq_init_state: slot_zero = 0xbac80000, is_master = 0
[    0.327300] Loading iSCSI transport class v2.0-870.
[    0.330096] usbcore: registered new interface driver smsc95xx
[    0.332355] dwc_otg: version 3.00a 10-AUG-2012 (platform bus)
[    0.534835] Core Release: 2.80a
[    0.536917] Setting default values for core params
[    0.539082] Finished setting default values for core params
[    0.741636] Using Buffer DMA mode
[    0.743795] Periodic Transfer Interrupt Enhancement - disabled
[    0.746028] Multiprocessor Interrupt Enhancement - disabled
[    0.748340] OTG VER PARAM: 0, OTG VER FLAG: 0
[    0.750598] Dedicated Tx FIFOs mode
[    0.753082] WARN::dwc_otg_hcd_init:1047: FIQ DMA bounce buffers: virt = 0xbac04000 dma = 0xfac04000 len=9024
[    0.757724] FIQ FSM acceleration enabled for :
Non-periodic Split Transactions
Periodic Split Transactions
High-Speed Isochronous Endpoints
Interrupt/Control Split Transaction hack enabled
[    0.769553] dwc_otg: Microframe scheduler enabled
[    0.769605] WARN::hcd_init_fiq:413: FIQ on core 1 at 0x8044844c
[    0.772042] WARN::hcd_init_fiq:414: FIQ ASM at 0x804487bc length 36
[    0.774414] WARN::hcd_init_fiq:439: MPHI regs_base at 0xbb87e000
[    0.776770] dwc_otg 3f980000.usb: DWC OTG Controller
[    0.779102] dwc_otg 3f980000.usb: new USB bus registered, assigned bus number 1
[    0.781465] dwc_otg 3f980000.usb: irq 62, io mem 0x00000000
[    0.783790] Init: Port Power? op_state=1
[    0.786048] Init: Power Port (0)
[    0.788454] usb usb1: New USB device found, idVendor=1d6b, idProduct=0002
[    0.790767] usb usb1: New USB device strings: Mfr=3, Product=2, SerialNumber=1
[    0.793072] usb usb1: Product: DWC OTG Controller
[    0.795323] usb usb1: Manufacturer: Linux 4.4.38-v7+ dwc_otg_hcd
[    0.797599] usb usb1: SerialNumber: 3f980000.usb
[    0.800663] hub 1-0:1.0: USB hub found
[    0.802871] hub 1-0:1.0: 1 port detected
[    0.805571] dwc_otg: FIQ enabled
[    0.805581] dwc_otg: NAK holdoff enabled
[    0.805589] dwc_otg: FIQ split-transaction FSM enabled
[    0.805622] Module dwc_common_port init
[    0.805875] usbcore: registered new interface driver usb-storage
[    0.808295] mousedev: PS/2 mouse device common for all mice
[    0.811166] bcm2835-cpufreq: min=600000 max=1200000
[    0.813627] sdhci: Secure Digital Host Controller Interface driver
[    0.815892] sdhci: Copyright(c) Pierre Ossman
[    0.818466] sdhost: log_buf @ bac07000 (fac07000)
[    0.878208] mmc0: sdhost-bcm2835 loaded - DMA enabled (>1)
[    0.882729] mmc-bcm2835 3f300000.mmc: mmc_debug:0 mmc_debug2:0
[    0.885024] mmc-bcm2835 3f300000.mmc: DMA channel allocated
[    0.935448] mmc0: host does not support reading read-only switch, assuming write-enable
[    0.938294] sdhci-pltfm: SDHCI platform and OF driver helper
[    0.938687] ledtrig-cpu: registered to indicate activity on CPUs
[    0.938803] hidraw: raw HID events driver (C) Jiri Kosina
[    0.938984] usbcore: registered new interface driver usbhid
[    0.938987] usbhid: USB HID core driver
[    0.939528] Initializing XFRM netlink socket
[    0.939554] NET: Registered protocol family 17
[    0.939686] Key type dns_resolver registered
[    0.940186] Registering SWP/SWPB emulation handler
[    0.940985] registered taskstats version 1
[    0.941163] vc-sm: Videocore shared memory driver
[    0.941172] [vc_sm_connected_init]: start
[    0.949878] [vc_sm_connected_init]: end - returning 0
[    0.951240] 3f201000.uart: ttyAMA0 at MMIO 0x3f201000 (irq = 87, base_baud = 0) is a PL011 rev2
[    0.951614] of_cfs_init
[    0.951691] of_cfs_init: OK
[    0.979125] Waiting for root device /dev/mmcblk0p2...
[    0.981006] mmc0: new high speed SDHC card at address 59b4
[    0.981599] mmcblk0: mmc0:59b4 00000 14.9 GiB
[    0.984282]  mmcblk0: p1 p2
[    0.998289] Indeed it is in host mode hprt0 = 00021501
[    1.002223] mmc1: queuing unknown CIS tuple 0x80 (2 bytes)
[    1.005799] mmc1: queuing unknown CIS tuple 0x80 (3 bytes)
[    1.009285] mmc1: queuing unknown CIS tuple 0x80 (3 bytes)
[    1.012103] EXT4-fs (mmcblk0p2): mounted filesystem with ordered data mode. Opts: (null)
[    1.013918] mmc1: queuing unknown CIS tuple 0x80 (7 bytes)
[    1.018047] VFS: Mounted root (ext4 filesystem) readonly on device 179:2.
[    1.028607] devtmpfs: mounted
[    1.031457] Freeing unused kernel memory: 476K (807eb000 - 80862000)
[    1.105242] mmc1: new high speed SDIO card at address 0001
[    1.178232] usb 1-1: new high-speed USB device number 2 using dwc_otg
[    1.180763] Indeed it is in host mode hprt0 = 00001101
[    1.273912] random: systemd: uninitialized urandom read (16 bytes read, 28 bits of entropy available)
[    1.282788] systemd[1]: systemd 215 running in system mode. (+PAM +AUDIT +SELINUX +IMA +SYSVINIT +LIBCRYPTSETUP +GCRYPT +ACL +XZ -SECCOMP -APPARMOR)
[    1.288457] systemd[1]: Detected architecture 'arm'.
[    1.378493] usb 1-1: New USB device found, idVendor=0424, idProduct=9514
[    1.381262] usb 1-1: New USB device strings: Mfr=0, Product=0, SerialNumber=0
[    1.384848] hub 1-1:1.0: USB hub found
[    1.387804] hub 1-1:1.0: 5 ports detected
[    1.427819] NET: Registered protocol family 10
[    1.432041] systemd[1]: Inserted module 'ipv6'
[    1.436430] systemd[1]: Set hostname to <raspberrypi>.
[    1.574933] random: systemd-sysv-ge: uninitialized urandom read (16 bytes read, 65 bits of entropy available)
[    1.668225] usb 1-1.1: new high-speed USB device number 3 using dwc_otg
[    1.730577] random: systemd: uninitialized urandom read (16 bytes read, 69 bits of entropy available)
[    1.737337] random: systemd: uninitialized urandom read (16 bytes read, 69 bits of entropy available)
[    1.744283] random: systemd: uninitialized urandom read (16 bytes read, 69 bits of entropy available)
[    1.766457] random: systemd: uninitialized urandom read (16 bytes read, 69 bits of entropy available)
[    1.772365] usb 1-1.1: New USB device found, idVendor=0424, idProduct=ec00
[    1.773020] random: systemd: uninitialized urandom read (16 bytes read, 69 bits of entropy available)
[    1.773179] random: systemd: uninitialized urandom read (16 bytes read, 69 bits of entropy available)
[    1.787423] usb 1-1.1: New USB device strings: Mfr=0, Product=0, SerialNumber=0
[    1.793670] smsc95xx v1.0.4
[    1.812824] random: systemd: uninitialized urandom read (16 bytes read, 71 bits of entropy available)
[    1.819947] random: systemd: uninitialized urandom read (16 bytes read, 71 bits of entropy available)
[    1.861828] smsc95xx 1-1.1:1.0 eth0: register 'smsc95xx' at usb-3f980000.usb-1.1, smsc95xx USB 2.0 Ethernet, b8:27:eb:cd:3a:54
[    1.964361] systemd[1]: Cannot add dependency job for unit regenerate_ssh_host_keys.service, ignoring: Unit regenerate_ssh_host_keys.service failed to load: No such file or directory.
[    1.978626] systemd[1]: Starting Forward Password Requests to Wall Directory Watch.
[    1.985923] systemd[1]: Started Forward Password Requests to Wall Directory Watch.
[    1.993053] systemd[1]: Starting Remote File Systems (Pre).
[    2.000493] systemd[1]: Reached target Remote File Systems (Pre).
[    2.004088] systemd[1]: Starting Arbitrary Executable File Formats File System Automount Point.
[    2.015032] systemd[1]: Set up automount Arbitrary Executable File Formats File System Automount Point.
[    2.021911] systemd[1]: Starting Encrypted Volumes.
[    2.029008] systemd[1]: Reached target Encrypted Volumes.
[    2.032343] systemd[1]: Starting Swap.
[    2.038872] systemd[1]: Reached target Swap.
[    2.041942] systemd[1]: Expecting device dev-mmcblk0p1.device...
[    2.048099] systemd[1]: Expecting device dev-sda1.device...
[    2.054186] systemd[1]: Starting Root Slice.
[    2.060000] systemd[1]: Created slice Root Slice.
[    2.062725] systemd[1]: Starting User and Session Slice.
[    2.068462] systemd[1]: Created slice User and Session Slice.
[    2.071158] systemd[1]: Starting Delayed Shutdown Socket.
[    2.076631] systemd[1]: Listening on Delayed Shutdown Socket.
[    2.079199] systemd[1]: Starting /dev/initctl Compatibility Named Pipe.
[    2.084549] systemd[1]: Listening on /dev/initctl Compatibility Named Pipe.
[    2.087144] systemd[1]: Starting Journal Socket (/dev/log).
[    2.092626] systemd[1]: Listening on Journal Socket (/dev/log).
[    2.095309] systemd[1]: Starting udev Control Socket.
[    2.100890] systemd[1]: Listening on udev Control Socket.
[    2.103563] systemd[1]: Starting udev Kernel Socket.
[    2.108213] usb 1-1.4: new high-speed USB device number 4 using dwc_otg
[    2.111750] systemd[1]: Listening on udev Kernel Socket.
[    2.114473] systemd[1]: Starting Journal Socket.
[    2.120139] systemd[1]: Listening on Journal Socket.
[    2.122907] systemd[1]: Starting System Slice.
[    2.128609] systemd[1]: Created slice System Slice.
[    2.131283] systemd[1]: Starting File System Check on Root Device...
[    2.178717] systemd[1]: Starting system-systemd\x2dfsck.slice.
[    2.184622] systemd[1]: Created slice system-systemd\x2dfsck.slice.
[    2.187441] systemd[1]: Starting system-autologin.slice.
[    2.193443] systemd[1]: Created slice system-autologin.slice.
[    2.196400] systemd[1]: Starting Increase datagram queue length...
[    2.204487] systemd[1]: Starting Restore / save the current clock...
[    2.209524] usb 1-1.4: New USB device found, idVendor=0781, idProduct=5583
[    2.209534] usb 1-1.4: New USB device strings: Mfr=1, Product=2, SerialNumber=3
[    2.209541] usb 1-1.4: Product: Ultra Fit
[    2.209548] usb 1-1.4: Manufacturer: SanDisk
[    2.209554] usb 1-1.4: SerialNumber: 4C530123080904121514
[    2.210573] usb-storage 1-1.4:1.0: USB Mass Storage device detected
[    2.211153] scsi host0: usb-storage 1-1.4:1.0
[    2.239175] systemd[1]: Starting Create list of required static device nodes for the current kernel...
[    2.250404] systemd[1]: Mounting POSIX Message Queue File System...
[    2.259616] systemd[1]: Mounting Debug File System...
[    2.293618] systemd[1]: Started Set Up Additional Binary Formats.
[    2.296701] systemd[1]: Mounted Huge Pages File System.
[    2.300298] systemd[1]: Starting udev Coldplug all Devices...
[    2.327828] systemd[1]: Starting Load Kernel Modules...
[    2.336268] systemd[1]: Starting Slices.
[    2.342622] systemd[1]: Reached target Slices.
[    2.354499] systemd[1]: Mounted Debug File System.
[    2.361292] systemd[1]: Mounted POSIX Message Queue File System.
[    2.367959] systemd[1]: Started Increase datagram queue length.
[    2.394079] systemd[1]: Started Restore / save the current clock.
[    2.398035] fuse init (API version 7.23)
[    2.404004] systemd[1]: Started Create list of required static device nodes for the current kernel.
[    2.422750] i2c /dev entries driver
[    2.432351] systemd[1]: Started Load Kernel Modules.
[    2.437068] systemd[1]: Time has been changed
[    2.463083] systemd[1]: Started udev Coldplug all Devices.
[    2.516413] systemd[1]: Started File System Check on Root Device.
[    2.596757] systemd[1]: Starting Apply Kernel Variables...
[    2.659141] systemd[1]: Mounting Configuration File System...
[    2.667806] systemd[1]: Mounting FUSE Control File System...
[    2.677876] systemd[1]: Starting Create Static Device Nodes in /dev...
[    2.684320] systemd[1]: Starting Syslog Socket.
[    2.688052] systemd[1]: Listening on Syslog Socket.
[    2.688329] systemd[1]: Starting Journal Service...
[    2.698080] systemd[1]: Started Journal Service.
[    2.882717] systemd-udevd[140]: starting version 215
[    3.220846] scsi 0:0:0:0: Direct-Access     SanDisk  Ultra Fit        1.00 PQ: 0 ANSI: 6
[    3.222689] sd 0:0:0:0: [sda] 62566488 512-byte logical blocks: (32.0 GB/29.8 GiB)
[    3.223687] sd 0:0:0:0: [sda] Write Protect is off
[    3.223715] sd 0:0:0:0: [sda] Mode Sense: 43 00 00 00
[    3.224189] sd 0:0:0:0: [sda] Write cache: disabled, read cache: enabled, doesn't support DPO or FUA
[    3.260881]  sda: sda1
[    3.263295] sd 0:0:0:0: [sda] Attached SCSI removable disk
[    3.318730] bcm2835-wdt 3f100000.watchdog: Broadcom BCM2835 watchdog timer
[    3.320873] gpiomem-bcm2835 3f200000.gpiomem: Initialised: Registers at 0x3f200000
[    3.466465] sd 0:0:0:0: Attached scsi generic sg0 type 0
[    3.570403] EXT4-fs (mmcblk0p2): re-mounted. Opts: (null)
[    3.607633] usbcore: registered new interface driver brcmfmac
[    3.697725] random: nonblocking pool is initialized
[    3.767721] brcmfmac: brcmf_c_preinit_dcmds: Firmware version = wl0: May 27 2016 00:13:38 version 7.45.41.26 (r640327) FWID 01-df77e4a7
[    3.799097] brcmfmac: brcmf_cfg80211_reg_notifier: not a ISO3166 code
[    4.012468] brcmfmac: brcmf_cfg80211_reg_notifier: not a ISO3166 code
[    4.012495] cfg80211: World regulatory domain updated:
[    4.012504] cfg80211:  DFS Master region: unset
[    4.012513] cfg80211:   (start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp), (dfs_cac_time)
[    4.012530] cfg80211:   (2402000 KHz - 2472000 KHz @ 40000 KHz), (N/A, 2000 mBm), (N/A)
[    4.012543] cfg80211:   (2457000 KHz - 2482000 KHz @ 40000 KHz), (N/A, 2000 mBm), (N/A)
[    4.012555] cfg80211:   (2474000 KHz - 2494000 KHz @ 20000 KHz), (N/A, 2000 mBm), (N/A)
[    4.012569] cfg80211:   (5170000 KHz - 5250000 KHz @ 80000 KHz, 160000 KHz AUTO), (N/A, 2000 mBm), (N/A)
[    4.012584] cfg80211:   (5250000 KHz - 5330000 KHz @ 80000 KHz, 160000 KHz AUTO), (N/A, 2000 mBm), (0 s)
[    4.012597] cfg80211:   (5490000 KHz - 5730000 KHz @ 160000 KHz), (N/A, 2000 mBm), (0 s)
[    4.012609] cfg80211:   (5735000 KHz - 5835000 KHz @ 80000 KHz), (N/A, 2000 mBm), (N/A)
[    4.012621] cfg80211:   (57240000 KHz - 63720000 KHz @ 2160000 KHz), (N/A, 0 mBm), (N/A)
[    4.182173] systemd-journald[138]: Received request to flush runtime journal from PID 1
[    4.934294] uart-pl011 3f201000.uart: no DMA platform data
[    5.208729] Adding 102396k swap on /var/swap.  Priority:-1 extents:3 across:208896k SSFS
[    5.331268] cfg80211: Regulatory domain changed to country: ES
[    5.331283] cfg80211:  DFS Master region: ETSI
[    5.331288] cfg80211:   (start_freq - end_freq @ bandwidth), (max_antenna_gain, max_eirp), (dfs_cac_time)
[    5.331296] cfg80211:   (2400000 KHz - 2483500 KHz @ 40000 KHz), (N/A, 2000 mBm), (N/A)
[    5.331304] cfg80211:   (5150000 KHz - 5250000 KHz @ 80000 KHz, 200000 KHz AUTO), (N/A, 2301 mBm), (N/A)
[    5.331311] cfg80211:   (5250000 KHz - 5350000 KHz @ 80000 KHz, 200000 KHz AUTO), (N/A, 2000 mBm), (0 s)
[    5.331318] cfg80211:   (5470000 KHz - 5725000 KHz @ 160000 KHz), (N/A, 2698 mBm), (0 s)
[    5.331324] cfg80211:   (57000000 KHz - 66000000 KHz @ 2160000 KHz), (N/A, 4000 mBm), (N/A)
[    5.549454] smsc95xx 1-1.1:1.0 eth0: hardware isn't capable of remote wakeup
[    5.549692] IPv6: ADDRCONF(NETDEV_UP): eth0: link is not ready
[    7.096523] IPv6: ADDRCONF(NETDEV_CHANGE): eth0: link becomes ready
[    7.097338] smsc95xx 1-1.1:1.0 eth0: link up, 100Mbps, full-duplex, lpa 0xC5E1
[    9.515485] Bluetooth: Core ver 2.21
[    9.515571] NET: Registered protocol family 31
[    9.515583] Bluetooth: HCI device and connection manager initialized
[    9.515608] Bluetooth: HCI socket layer initialized
[    9.515625] Bluetooth: L2CAP socket layer initialized
[    9.515660] Bluetooth: SCO socket layer initialized
[    9.522843] Bluetooth: HCI UART driver ver 2.3
[    9.522867] Bluetooth: HCI UART protocol H4 registered
[    9.522877] Bluetooth: HCI UART protocol Three-wire (H5) registered
[    9.523040] Bluetooth: HCI UART protocol BCM registered
[    9.741047] Bluetooth: BNEP (Ethernet Emulation) ver 1.3
[    9.741059] Bluetooth: BNEP filters: protocol multicast
[    9.741072] Bluetooth: BNEP socket layer initialized
[   13.796334] PPP generic driver version 2.4.2
[   13.887817] NET: Registered protocol family 24
pi@raspberrypi:~ $ 
This line means it's using swap? I've got like 750MB free RAM

Code: Select all

[    5.208729] Adding 102396k swap on /var/swap.  Priority:-1 extents:3 across:208896k SSFS

EnfermeraSexy
Posts: 17
Joined: Mon Jan 02, 2017 6:19 pm

Re: Stuttering when writing to a USB

Tue Jan 03, 2017 2:51 pm

Here you can see what happens when i read/write something from NTFS/EXT4 and the CPU usage

Writing to NTFS
Image
Reading from NTFS
Image

Writing to EXT4
Image
Reading from EXT4
Image

ejolson
Posts: 5137
Joined: Tue Mar 18, 2014 11:47 am

Re: Stuttering when writing to a USB

Tue Jan 03, 2017 5:08 pm

EnfermeraSexy wrote:What i'm looking for? I don't really know how does Linux work, i'll attach the log. I was downloading something like for 4 hours
You would look for things like "I/O error" or "lost interrupt" or "resetting..." in the log file. No such messages appear, in fact, no messages past the initial boot messages are there. Note that every time you reboot, the dmesg queue gets cleared and started over, so only type dmesg after the system has been up for a while and misbehaving. Assuming you've been experiencing the stuttering immediately prior to typing dmesg, the unwanted behavior seems not caused by an obvious hardware fault.

ejolson
Posts: 5137
Joined: Tue Mar 18, 2014 11:47 am

Re: Stuttering when writing to a USB

Tue Jan 03, 2017 7:11 pm

EnfermeraSexy wrote:Here you can see what happens when i read/write something from NTFS/EXT4 and the CPU usage.
Most people would consider stuttering to be exactly what is expected when using a toy computer as a file server. However, your discovery that NTFS doesn't stutter in the same way is not only new to me but surprising.

Since the external disk and Ethernet share a single USB channel, the Pi may drop Ethernet packets while writing large amounts of data. Since ext4 doesn't write data until the buffer cache fills and then writes a lot, ext4 may cause more dropped packets than NTFS. You may be able to see a difference in the dropped packet statistics reported by ifconfig. There are a number of parameters that may tune the performance of the I/O scheduler and ext4 filesystem in a way more suitable for the Pi. Maybe setting the I/O scheduler to none and the associated parameters to reduce the tendency of the scheduler bunch requests together would help.

EnfermeraSexy
Posts: 17
Joined: Mon Jan 02, 2017 6:19 pm

Re: Stuttering when writing to a USB

Tue Jan 03, 2017 7:40 pm

ejolson wrote:You would look for things like "I/O error" or "lost interrupt" or "resetting..." in the log file. No such messages appear, in fact, no messages past the initial boot messages are there. Note that every time you reboot, the dmesg queue gets cleared and started over, so only type dmesg after the system has been up for a while and misbehaving. Assuming you've been experiencing the stuttering immediately prior to typing dmesg, the unwanted behavior seems not caused by an obvious hardware fault.
PI was running for over 4-5 hours and writed like 30 GB of data before i copy that log, so we can say that there's no problem with my hardware.

EnfermeraSexy
Posts: 17
Joined: Mon Jan 02, 2017 6:19 pm

Re: Stuttering when writing to a USB

Tue Jan 03, 2017 10:54 pm

ejolson wrote:Most people would consider stuttering to be exactly what is expected when using a toy computer as a file server. However, your discovery that NTFS doesn't stutter in the same way is not only new to me but surprising.

Since the external disk and Ethernet share a single USB channel, the Pi may drop Ethernet packets while writing large amounts of data. Since ext4 doesn't write data until the buffer cache fills and then writes a lot, ext4 may cause more dropped packets than NTFS. You may be able to see a difference in the dropped packet statistics reported by ifconfig. There are a number of parameters that may tune the performance of the I/O scheduler and ext4 filesystem in a way more suitable for the Pi. Maybe setting the I/O scheduler to none and the associated parameters to reduce the tendency of the scheduler bunch requests together would help.

Code: Select all

eth0      Link encap:Ethernet  HWaddr b8:27:eb:cd:3a:54  
          inet addr:192.168.1.3  Bcast:192.168.1.255  Mask:255.255.255.0
          inet6 addr: fe80::ce55:20a5:8cbc:a507/64 Scope:Link
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:15953087 errors:0 dropped:2 overruns:0 frame:0
          TX packets:7483337 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:1000 
          RX bytes:2226462519 (2.0 GiB)  TX bytes:531896109 (507.2 MiB)
I don't think i'm getting errors through ethernet. I tried parameters from your links, but nothing change... i think it should perform better... i mean, it's not using all his resources... low CPU usage... and it has a really powerful hardware attached... i didn't expected so much but well... seems to be unstable.

Thanks both for your time.

ejolson
Posts: 5137
Joined: Tue Mar 18, 2014 11:47 am

Re: Stuttering when writing to a USB

Wed Jan 04, 2017 6:49 am

EnfermeraSexy wrote:I don't think i'm getting errors through ethernet. I tried parameters from your links, but nothing change... i think it should perform better... i mean, it's not using all his resources... low CPU usage... and it has a really powerful hardware attached... i didn't expected so much but well... seems to be unstable.
It seems I'm not able to figure out what's going on. Maybe someone else on this forum has seen this stuttering behavior and knows how to fix it.

Just for the record, my understanding is that

1. You are using a Pi 3B to serve files from a USB drive over Samba to a Windows machine.

2. Keeping all hardware the same, the system works better when the USB drive is formatted using NTFS but stutters when formatted using ext4.

Further details about your setup might help others to reproduce and diagnose the stuttering. I find your results very interesting. Could you please update this thread if I've misunderstood the issue or you discover anything more. All the best!

EnfermeraSexy
Posts: 17
Joined: Mon Jan 02, 2017 6:19 pm

Re: Stuttering when writing to a USB

Wed Jan 04, 2017 4:12 pm

ejolson wrote:It seems I'm not able to figure out what's going on. Maybe someone else on this forum has seen this stuttering behavior and knows how to fix it.

Just for the record, my understanding is that

1. You are using a Pi 3B to serve files from a USB drive over Samba to a Windows machine.

2. Keeping all hardware the same, the system works better when the USB drive is formatted using NTFS but stutters when formatted using ext4.

Further details about your setup might help others to reproduce and diagnose the stuttering. I find your results very interesting. Could you please update this thread if I've misunderstood the issue or you discover anything more. All the best!
Not at all. In log it says i've got a Raspberry Pi 3 Model B Rev 1.2, if that helps, but it stutters in every filesystem. What happens it's that PI handles better NTFS, but it doesn't mean it doesn't stutter. In my test i was writing just a single file, but if i use for example Transmission (my maximum speed connection it's 1.8 MB/s, not too high) it will stutter on every system, just less in NTFS or FAT32.
EXT4 it's unusable if you want to use Transmission or any single tool, and i can see how transmission got freezes for like 1 or 2 minutes after working for a couple of minutes if i use EXT4.

It stutters too if i try to use SD card instead of USB, so it's stuttering in every memory. I don't get why is this happening, i tried a lot of different configs, but it doesn't matter.

ejolson
Posts: 5137
Joined: Tue Mar 18, 2014 11:47 am

Re: Stuttering when writing to a USB

Wed Jan 04, 2017 6:56 pm

EnfermeraSexy wrote:In my test i was writing just a single file, but if i use for example Transmission (my maximum speed connection it's 1.8 MB/s, not too high) it will stutter on every system, just less in NTFS or FAT32.
A speed of 1.8 MB/s sounds like a slow residential Internet connection, while the graphs were clearly made using a LAN. Could you describe in more details how the graphs were made?

EnfermeraSexy
Posts: 17
Joined: Mon Jan 02, 2017 6:19 pm

Re: Stuttering when writing to a USB

Wed Jan 04, 2017 7:25 pm

ejolson wrote:A speed of 1.8 MB/s sounds like a slow residential Internet connection, while the graphs were clearly made using a LAN. Could you describe in more details how the graphs were made?
I told that because Transmission (as every torrent client) has more threats reading and writing to the drive at the same time, so it makes more stuttering than being writing a single file with one thread.

The graphs were made LAN to LAN, just to explain what my problem is and how PI can't make properly a really easy task.

Heater
Posts: 15808
Joined: Tue Jul 17, 2012 3:02 pm

Re: Stuttering when writing to a USB

Wed Jan 04, 2017 7:40 pm

I think we need to isolate some parts here. We have network I/O, file system read/write, threads or not....

Is this stuttering thing a file system phenomena? Is it the network? Is it the threads? A combination of any two or all of those?

Presumably some simple programs could perform these actions in isolation and report any stuttering.

I'm curious because when my Pi seem to "hang" for seconds end when I SSH into them. I run them headless. I have always wondered why.
Memory in C++ is a leaky abstraction .

EnfermeraSexy
Posts: 17
Joined: Mon Jan 02, 2017 6:19 pm

Re: Stuttering when writing to a USB

Wed Jan 04, 2017 7:53 pm

Heater wrote:I think we need to isolate some parts here. We have network I/O, file system read/write, threads or not....

Is this stuttering thing a file system phenomena? Is it the network? Is it the threads? A combination of any two or all of those?

Presumably some simple programs could perform these actions in isolation and report any stuttering.

I'm curious because when my Pi seem to "hang" for seconds end when I SSH into them. I run them headless. I have always wondered why.
If you tell me some of that programs i can do some test, i'm new in linux and i don't know so much about it.

Let's see. The real problem are the system read/writes. The more threads you use, the more it stutters and the worse it perform. It increase if you use a Linux filesystem, like EXT4. Doesn't matter if you do it in local (not in local network) or you use the network, it's just to make things clearly with some screenshot.

ejolson
Posts: 5137
Joined: Tue Mar 18, 2014 11:47 am

Re: Stuttering when writing to a USB

Wed Jan 04, 2017 8:26 pm

Heater wrote:.I'm curious because when my Pi seem to "hang" for seconds end when I SSH into them. I run them headless. I have always wondered why.
The delay in ssh is probably caused by timing out on a reverse DNS lookup. My theory is the stuttering here is related to lost Ethernet packets and the resulting exponential back-off of retransmit timeouts at the TCP/IP level.

Heater
Posts: 15808
Joined: Tue Jul 17, 2012 3:02 pm

Re: Stuttering when writing to a USB

Wed Jan 04, 2017 8:44 pm

ejolson,

Not sure what you mean. I'm talking about delays after I'm logged in and busy typing commands. Even if I don't expect those commands to be doing anything very disk or network heavy.
Memory in C++ is a leaky abstraction .

EnfermeraSexy
Posts: 17
Joined: Mon Jan 02, 2017 6:19 pm

Re: Stuttering when writing to a USB

Wed Jan 04, 2017 8:58 pm

ejolson wrote:
Heater wrote:.I'm curious because when my Pi seem to "hang" for seconds end when I SSH into them. I run them headless. I have always wondered why.
The delay in ssh is probably caused by timing out on a reverse DNS lookup. My theory is the stuttering here is related to lost Ethernet packets and the resulting exponential back-off of retransmit timeouts at the TCP/IP level.
It happens too if you copy something from one USB to another.

ejolson
Posts: 5137
Joined: Tue Mar 18, 2014 11:47 am

Re: Stuttering when writing to a USB

Wed Jan 04, 2017 9:12 pm

Heater wrote:ejolson,

Not sure what you mean. I'm talking about delays after I'm logged in and busy typing commands. Even if I don't expect those commands to be doing anything very disk or network heavy.
Aha! Never mind then.

ejolson
Posts: 5137
Joined: Tue Mar 18, 2014 11:47 am

Re: Stuttering when writing to a USB

Wed Jan 04, 2017 9:33 pm

EnfermeraSexy wrote:It happens too if you copy something from one USB to another.
That is an interesting claim and could certainly help discover the cause of the stuttering.

Sorry that I'm currently traveling and can't try any experiments myself. Suppose I format two USB drives with ext4 and then time how long it takes to copy a 6GB file from one to the other. Next, I reformat the same drives with NTFS and time how long it takes to copy the same file from one to the other. As ext4 is a native filesystem while NTFS runs through an emulation layer, common sense dictates the ext4 to ext4 copy should be faster. Is this not actually the case?

What happens if you open an additional terminal window and run the infinite loop

$ while sync; do sync; done

while performing the tests? Does the stuttering get worse or better?

Heater
Posts: 15808
Joined: Tue Jul 17, 2012 3:02 pm

Re: Stuttering when writing to a USB

Wed Jan 04, 2017 9:36 pm

What emulation layer for NTFS? As far as I know it's a regular Linux driver modules. Same like EXT etc.
Memory in C++ is a leaky abstraction .

ejolson
Posts: 5137
Joined: Tue Mar 18, 2014 11:47 am

Re: Stuttering when writing to a USB

Wed Jan 04, 2017 9:42 pm

Heater wrote:What emulation layer for NTFS? As far as I know it's a regular Linux driver modules. Same like EXT etc.
I think NTFS works through FUSE.

Heater
Posts: 15808
Joined: Tue Jul 17, 2012 3:02 pm

Re: Stuttering when writing to a USB

Wed Jan 04, 2017 9:47 pm

Ha, OK. Still FUSE is not an emulation. But I guess it might have overheads of it's own.

Aside: I love Debian's description of NTFS:

"NTFS (Windows NT FileSystem) - the latest in a long line of minor, proprietary, non-portable, file system formats promulgated by a commercial software vendor based in the Pacific North West, USA. Closed technology."
Memory in C++ is a leaky abstraction .

EnfermeraSexy
Posts: 17
Joined: Mon Jan 02, 2017 6:19 pm

Re: Stuttering when writing to a USB

Thu Jan 05, 2017 12:32 pm

ejolson wrote:That is an interesting claim and could certainly help discover the cause of the stuttering.

Sorry that I'm currently traveling and can't try any experiments myself. Suppose I format two USB drives with ext4 and then time how long it takes to copy a 6GB file from one to the other. Next, I reformat the same drives with NTFS and time how long it takes to copy the same file from one to the other. As ext4 is a native filesystem while NTFS runs through an emulation layer, common sense dictates the ext4 to ext4 copy should be faster. Is this not actually the case?

What happens if you open an additional terminal window and run the infinite loop

$ while sync; do sync; done

while performing the tests? Does the stuttering get worse or better?
Ok here are my test, with the .img from raspbian (4GB):
From NTFS to NTFS took 6 minutes and 54 seconds
From EXT4 to EXT4 took 7 minutes and 17 seconds

If i use your command seems to be worse, not too much but i think it stutter more.

ejolson
Posts: 5137
Joined: Tue Mar 18, 2014 11:47 am

Re: Stuttering when writing to a USB

Fri Jan 06, 2017 10:37 pm

EnfermeraSexy wrote:Ok here are my test, with the .img from raspbian (4GB):
From NTFS to NTFS took 6 minutes and 54 seconds
From EXT4 to EXT4 took 7 minutes and 17 seconds
It looks like it took about 5% longer with EXT4 than NTFS. While significant, the difference is not nearly as noticeable as the differences demonstrated by your earlier graphs when networking is involved. In particular a 5% slowdown could be explained by differences in the way the journals are updated and doesn't clearly indicate something is broken in the same way the graphs do.

To diagnose what is happening in the graphs someone else needs to replicate the behavior. It will be a couple weeks before I can try to do this but maybe someone else can obtain similar results sooner. Again, specific details how the graphs were obtained would be useful. Were these graphs obtained when serving files to a Macintosh computer? What kind of external USB disks are you using?

EnfermeraSexy
Posts: 17
Joined: Mon Jan 02, 2017 6:19 pm

Re: Stuttering when writing to a USB

Sat Jan 07, 2017 2:36 pm

ejolson wrote: It looks like it took about 5% longer with EXT4 than NTFS. While significant, the difference is not nearly as noticeable as the differences demonstrated by your earlier graphs when networking is involved. In particular a 5% slowdown could be explained by differences in the way the journals are updated and doesn't clearly indicate something is broken in the same way the graphs do.

To diagnose what is happening in the graphs someone else needs to replicate the behavior. It will be a couple weeks before I can try to do this but maybe someone else can obtain similar results sooner. Again, specific details how the graphs were obtained would be useful. Were these graphs obtained when serving files to a Macintosh computer? What kind of external USB disks are you using?
Using system-config-samba in rpi, sharing the folder from my USB to a machine with Windows 10. The USB it's a Sandisk Ultra Fit, you can see it in amazon
https://www.amazon.es/Memoria-flash-San ... ER2CS?th=1
Tried an external HDD, and nothing changes. Both (rpi and my computer) are connected through ethernet (CAT.6) to a personal router (ASUS n-17u).

Please, don't look into the graphs, that was just an example, try transmission, and you will see that is useless to use it in rpi, it will always stutter, you can try any config you want, but it won't work better. No matters if you use NTFS or EXT4.

Return to “General discussion”