exterrestris
-
Posts
5 -
Joined
-
Last visited
-
Days Won
1
Posts posted by exterrestris
-
-
DrivePool doesn't care about the drives in the pool having drive letters, so you'll be fine there. SnapRAID on the other hand needs to be able to access the drives - if you don't want drive letters for each, then you can mount each drive in an empty folder instead (as long as the drive is NTFS).
I use the following folder structure:
C:\Snapraid \Data \1 \2 \3 \4 \Parity \1 \Cache \1 \2Each of the numbered folders is a mount point for a disk - SnapRAID is then configured to point at each of the data drives and the parity drive. Best practice is to set the "root" of each data disk in the SnapRAID config to the hidden Drivepool poolpart directory (because if/when you swap out a disk, Drivepool will assign a new ID to the poolpart directory, so the "root" would change). Also make sure that Drivepool isn't going to do any balancing between the data drives, as that will just cause an excessive number of changes each time SnapRAID runs.
Since SnapRAID doesn't like it when you write to the disk whilst it's doing a sync/scrub, I also use two cache drives with the SSD Optimizer plugin. These are not protected by SnapRAID, but are instead set up as another Drivepool with duplication enabled, which is then part of the main pool, like so:
SnapRAID (Pool) - Mounted as E:\ - C:\Snapraid\Data\1 - C:\Snapraid\Data\2 - C:\Snapraid\Data\3 - C:\Snapraid\Data\4 - Data Cache (Pool) - C:\Snapraid\Cache\1 - C:\Snapraid\Cache\2This means that new files are always written to the cache drives, and will not interfere with SnapRAID if it happens to be running, but those files still have protection against disk failure. The SSD Optimizer is the only balancer enabled on the main pool (well, except for the Scanner drive evacuation balancer), and balances into the main pool once a day, so the cache disks only need to be big enough to hold the largest file that is likely to be written to the pool.
-
I'm running StableBit Scanner 2.5.4.3216. I've also had to change the driver I'm using back to the default Windows (non-UASP) driver, as the ASUS UAS Storage Driver seems to be a bit buggy, but this hasn't changed the lack of health data. With the ASUS UAS Storage Driver, attempting to do a surface scan causes the drive to stop responding after completing ~5% (read speed drops to 0 in Task Manager, but the disk remains 100% active), and it'll never pass that point even after stopping and restarting the scan. Switching back to using the default Windows driver (USB Mass Storage Device) allows the entire disk to be scanned without problems.
-
I have an Icy Box IB-1816M-C31 USB 3.1 Gen 2 enclosure containing a 512GB Samsung 950 Pro NVMe SSD, for which Scanner cannot read the SMART data. This enclosure uses a JMicron JMS583 USB-NVMe bridge, and shows up as a UASP device using the ASUS UAS Storage Driver (removing this driver and using the default driver, which takes it out of UASP mode, doesn't change anything).
CystalDiskInfo is able to read the SMART data, as is smartctl when run with the -d sntjmicron option to correctly identify the USB device.
-d sntjmicronC:\Program Files\smartmontools\bin>smartctl -a -d sntjmicron /dev/sdg smartctl 7.0 2018-12-30 r4883 [x86_64-w64-mingw32-w10-1803] (sf-7.0-1) Copyright (C) 2002-18, Bruce Allen, Christian Franke, www.smartmontools.org === START OF INFORMATION SECTION === Model Number: Samsung SSD 950 PRO 512GB Serial Number: S2GMNXAH113261W Firmware Version: 2B0QBXX7 PCI Vendor/Subsystem ID: 0x144d IEEE OUI Identifier: 0x002538 Controller ID: 1 Number of Namespaces: 1 Namespace 1 Size/Capacity: 512,110,190,592 [512 GB] Namespace 1 Utilization: 237,899,358,208 [237 GB] Namespace 1 Formatted LBA Size: 512 Namespace 1 IEEE EUI-64: 002538 5161b033cd Local Time is: Sat Jul 20 17:41:36 2019 GMTST Firmware Updates (0x06): 3 Slots Optional Admin Commands (0x0007): Security Format Frmw_DL Optional NVM Commands (0x001f): Comp Wr_Unc DS_Mngmt Wr_Zero Sav/Sel_Feat Maximum Data Transfer Size: 32 Pages Supported Power States St Op Max Active Idle RL RT WL WT Ent_Lat Ex_Lat 0 + 6.50W - - 0 0 0 0 5 5 1 + 5.80W - - 1 1 1 1 30 30 2 + 3.60W - - 2 2 2 2 100 100 3 - 0.0700W - - 3 3 3 3 500 5000 4 - 0.0050W - - 4 4 4 4 2000 22000 Supported LBA Sizes (NSID 0x1) Id Fmt Data Metadt Rel_Perf 0 + 512 0 0 === START OF SMART DATA SECTION === SMART overall-health self-assessment test result: PASSED SMART/Health Information (NVMe Log 0x02) Critical Warning: 0x00 Temperature: 51 Celsius Available Spare: 100% Available Spare Threshold: 10% Percentage Used: 2% Data Units Read: 40,017,269 [20.4 TB] Data Units Written: 74,499,658 [38.1 TB] Host Read Commands: 586,596,545 Host Write Commands: 842,689,794 Controller Busy Time: 3,890 Power Cycles: 3,841 Power On Hours: 10,578 Unsafe Shutdowns: 66 Media and Data Integrity Errors: 0 Error Information Log Entries: 13,386 Error Information (NVMe Log 0x01, max 64 entries) Num ErrCount SQId CmdId Status PELoc LBA NSID VS 0 13386 0 0x0b0e 0x4004 0x000 0 0 - 1 13385 0 0x0609 0x4004 0x000 0 0 - 2 13384 0 0x0508 0x4004 0x000 0 0 - 3 13383 0 0x0407 0x4004 0x000 0 0 - 4 13382 0 0x0306 0x4004 0x000 0 0 - 5 13381 0 0x0a06 0x4004 0x000 0 0 - 6 13380 0 0x010a 0x4004 0x000 0 0 - 7 13379 0 0x0e06 0x4004 0x000 0 0 - 8 13378 0 0x0d01 0x4004 0x000 0 0 - 9 13377 0 0x0c00 0x4004 0x000 0 0 - 10 13376 0 0x0b0f 0x4004 0x000 0 0 - 11 13375 0 0x0a0e 0x4004 0x000 0 0 - 12 13374 0 0x090d 0x4004 0x000 0 0 - 13 13373 0 0x080c 0x4004 0x000 0 0 - 14 13372 0 0x070b 0x4004 0x000 0 0 - 15 13371 0 0x060a 0x4004 0x000 0 0 - ... (48 entries not shown)
The stickied Direct I/O Test tool cannot read the SMART data either - it can only retrieve the following SSD data, no matter what method is used:
Version: 40 Model: SamsungSSD 950 PRO 2B0Q Serial number: DD56419883A90 Bus type: Usb Command queuing: True Device type: 0x00 Raw device properties: 0 Removable media: False Vendor ID: Samsung Product ID: SSD 950 PRO 2B0Q Product revision: 0204
Every other option shows as a red cross, unless ScsiPassthroughJmicron is selected. In that case, SMART status under Direct I/O becomes available (but not any other option), however it reports as Status: Failing
-
I'm using Drivepool (without duplication) in combination with Snapraid. This is working well, but there is a issue I would like to solve - Snapraid doesn't tolerate writing to the disk while it's running a sync (creating a snapshot). Most of the time this isn't a problem, as my daily sync is scheduled late in the morning after any backups or similar would usually have completed. However, periodically the sync will overlap with other activity causing it to fail - for example, when a full backup is being written (and so takes longer), or the sync includes a full backup (which takes longer to calculate parity).
What I thought I could do is add another disk and use the SSD Optimizer plugin to create a landing zone. This would be in the pool, but not protected by Snapraid, which would allow for writes to happen without interfering with Snapraid if a sync is running. Now most of the time I'd want this to balance pretty quickly, but only if a sync isn't running. What would be ideal was if there a way to enable/disable balancing from the command line - then the script I use to run my daily sync could disable balancing as it starts, and re-enable it once complete. Is this possible?
Drivepool With Snapraid
in General
Posted
My snapraid.conf is pretty standard - I haven't really changed any of the defaults (so I haven't included them). I choose to keep a copy of the content file on every disk, but that's not strictly necessary.
As for DrivePool balancers, yes, turn them all off. The Scanner is useful to keep if you want automatic evacuation of a failing drive, but not essential, and the SSD Optimiser is only necessary if you have a cache drive to use as a landing zone. If you don't use a landing zone, then you can disable automatic balancing, but if you do then you need it to balance periodically - once a day rather than immediately is best, as you ideally want the SnapRAID sync to happen shortly after the balance completes.
I'm not sure what the default behaviour of DrivePool is supposed to be when all balancers are disabled, but I think it does split evenly across the disks.