SSD speed in Pogoplug Pro

Hi all,

I have 2 pogoplug pro's running Debian bookworm, latest kernel:

Linux pogo01 5.15.158-oxnas-tld-2 #2 SMP PREEMPT Sun May 12 18:45:05 PDT 2024 armv6l GNU/Linux

In both plugs I have installed an SSD -> https://blog.westelinck.net/2023/01/adding-sata-ssd-to-pogoplug-pro.html

One of the plugs is now running pi-hole v6 and the v6 version is using a lot more memory than the v5 version ... and I see performance problems.

I am not an export, but I think it boils down to disk I/O ... when writing a 512MB file, these are the results:

dd if=/dev/zero of=/root/tempfile bs=1M count=512 conv=fdatasync
512+0 records in
512+0 records out
536870912 bytes (537 MB, 512 MiB) copied, 41.7972 s, 12.8 MB/s

So not that fast. I can also observe this in dmesg:

    7.887441] ata1: SATA link up 1.5 Gbps (SStatus 113 SControl 310)
[    7.893839] ehci-pci: EHCI PCI platform driver
[    7.901258] ata1.00: ATA-9: SanDisk SSD U100 16GB, 10.52.01, max UDMA/133
[    7.913331] ata1.00: 31277232 sectors, multi 1: LBA48 NCQ (depth 0/32)
[    7.921221] ata1.00: configured for UDMA/133
[    7.926192] sata_oxnas: resetting SATA core
[    8.107491] oxnas-ehci 40200100.ehci: EHCI Host Controller
[    8.113710] oxnas-ehci 40200100.ehci: new USB bus registered, assigned bus number 1
[    8.122419] oxnas-ehci 40200100.ehci: irq 60, io mem 0x40200100
[    8.157475] oxnas-ehci 40200100.ehci: USB 2.0 started, EHCI 1.00
[    8.164590] usb usb1: New USB device found, idVendor=1d6b, idProduct=0002, bcdDevice= 5.15
[    8.173591] usb usb1: New USB device strings: Mfr=3, Product=2, SerialNumber=1
[    8.181519] usb usb1: Product: EHCI Host Controller
[    8.187074] usb usb1: Manufacturer: Linux 5.15.158-oxnas-tld-2 ehci_hcd
[    8.194383] usb usb1: SerialNumber: 40200100.ehci
[    8.200885] hub 1-0:1.0: USB hub found
[    8.205404] hub 1-0:1.0: 2 ports detected
[    8.211566] usbcore: registered new interface driver cdc_wdm
[    8.218152] usbcore: registered new interface driver usb-storage
[    8.225041] usbcore: registered new interface driver usbserial_generic
[    8.232337] usbserial: USB Serial support registered for generic
[    8.239787] mousedev: PS/2 mouse device common for all mice
[    8.246259] i2c_dev: i2c /dev entries driver
[    8.253790] ledtrig-cpu: registered to indicate activity on CPUs
[    8.260723] hid: raw HID events driver (C) Jiri Kosina
[    8.266759] usbcore: registered new interface driver usbhid
[    8.273054] usbhid: USB HID core driver
[    8.279464] NET: Registered PF_INET6 protocol family
[    8.287823] Segment Routing with IPv6
[    8.292180] RPL Segment Routing with IPv6
[    8.296952] In-situ OAM (IOAM) with IPv6
[    8.301790] sit: IPv6, IPv4 and MPLS over IPv4 tunneling driver
[    8.310827] NET: Registered PF_PACKET protocol family
[    8.317171] Key type dns_resolver registered
[    8.323548] registered taskstats version 1
[    8.328465] Loading compiled-in X.509 certificates
[    8.334809] zswap: loaded using pool lzo/zbud
[    8.341447] Key type .fscrypt registered
[    8.346063] Key type fscrypt-provisioning registered
[    8.355182] Key type big_key registered
[    8.415912] Key type encrypted registered
[    8.507425] usb 1-1: new high-speed USB device number 2 using oxnas-ehci
[    8.667445] oxnas-pcie 47c00000.pcie: PCIe version/deviceID 0x82510b5
[    8.729259] usb 1-1: New USB device found, idVendor=05e3, idProduct=0608, bcdDevice=77.64
[    8.747397] usb 1-1: New USB device strings: Mfr=0, Product=1, SerialNumber=0
[    8.755210] usb 1-1: Product: USB2.0 Hub
[    8.769016] hub 1-1:1.0: USB hub found
[    8.773916] hub 1-1:1.0: 4 ports detected
[    9.617408] ata1: SATA link up 1.5 Gbps (SStatus 113 SControl 310)
[    9.628026] ata1.00: configured for UDMA/133
[    9.633011] sata_oxnas: resetting SATA core
[    9.677383] oxnas-pcie 47c00000.pcie: link down
[    9.683000] oxnas-pcie 47c00000.pcie: PCI host bridge to bus 0000:00
[    9.690086] pci_bus 0000:00: root bus resource [mem 0x48000000-0x49ffffff]
[    9.697697] pci_bus 0000:00: root bus resource [mem 0x4a000000-0x4bdfffff pref]
[    9.705681] pci_bus 0000:00: root bus resource [io  0x0000-0xfffff]
[    9.712660] pci_bus 0000:00: root bus resource [bus 00-7f]
[    9.719680] PCI: bus0: Fast back to back transfers enabled
[    9.726709] clk: Disabling unused clocks
[   10.417448] ata1.00: limiting speed to UDMA/100:PIO4
[   11.357432] ata1: SATA link up 1.5 Gbps (SStatus 113 SControl 310)
[   11.364981] ata1.00: configured for UDMA/100
[   11.370004] sata_oxnas: resetting SATA core
[   12.147445] ata1.00: limiting speed to UDMA/33:PIO4
[   13.087439] ata1: SATA link up 1.5 Gbps (SStatus 113 SControl 310)
[   13.094917] ata1.00: configured for UDMA/33
[   13.099855] sata_oxnas: resetting SATA core
[   14.817436] ata1: SATA link up 1.5 Gbps (SStatus 113 SControl 310)
[   14.824929] ata1.00: configured for UDMA/33
[   14.829864] ata1: EH pending after 5 tries, giving up

What does this mean? And why is it falling back to UDMA/33.

kennywest,

The SATA controller in this Pogo Pro has some limitation. So whatever you use as enclosure/adapter, it might be just that it does not play well with that specific controller.

And if you use a USB port for 5v power, make sure that it provides enough voltage.

-bodhi
===========================
Forum Wiki
bodhi's corner (buy bodhi a beer)

I saw some similar-ish issues between either 4.19 and 5.10 or 5.10 and 6.1 which I was able to track back to trim/discard being enabled by default for EXT4.

Some SATA port expanders, IDE->SATA converters and USB->SATA converters (and probably just older IDE/SATA controllers) will throw a fit if they receive an a TRIM command. I assume this is because their internal firmware predates that being added to the ATA spec but I've never dug that deep into it. It seems the kernel queries the drive for it's capabilities but doesn't know about other chips in between.

My main experience is with NAS's that use the Marvell 88SM4140 port expander which I use for testing with similar SSDs to the one you're using. In that case the devices weren't just slow but became unstable due to this issue.

You might be able to verify if this is happening by using fstrim and see if that triggers a bunch of errors. If that is the issue adding the nodiscard mount option should fix it.