jbartlett

Actually, I do create a bus tree so I can test each level for a better view.

This seems to highlight a recent thought of adding a system wide benchmark of the drives, reaching multiple controllers together and all together, with all combinations of controllers.

Install the DiskSpeed docker app in my sig, it will show/tell you if your controller is saturated.

I had eight WD 6TB Red Pros on a SAS PCIe 2 controller and I had to take two of them off of it before I could read all drives at the same time at the same speed as each drive on it's own. My UTT scores changed a large amount after.

Try my DiskSpeed (link in my sig) plugin's controller bandwidth test. It can help you determine if you've saturated your controller's capabilities.

If you're like me and can't run in Safe/Maint Mode, disabling network shares (globally or per share) will prevent people & apps from accessing the box from other machines but still let KVM & Docker run. In my case, I run my home web server through Docker and app/email server through a VM but neither is hitting against the array.

My backup server has no Parity drive. The Check button then does a Read check of all the array drives looking for errors.

Bingo! I didn't make the connection until reading the posts just prior to this.

Information attached. info.zip

Normal run from Beta 2. I had reset my parity settings to defaults after moving two drives off of the PCIe2 controller so it wasn't being saturated. Looks like defaults work for me. NormalSyncTestReport_2019_08_13_0102.txt Short run from Beta 3 ShortSyncTestReport_2019_08_13_1046.txt

I had but only the nvme drives so it likely was easy to miss. Here's the full report [0:0:0:0] disk usb:1-4:1.0 /dev/sda 32.0GB [1:0:0:0] disk sata:5000cca255c167b1 /dev/sdh 6.00TB [2:0:0:0] disk sata:5000cca24dce4c79 /dev/sdi 6.00TB [3:0:0:0] disk sata:5002538e40270cde /dev/sdj 1.00TB [4:0:0:0] disk sata:5002538e40270f85 /dev/sdk 1.00TB [5:0:0:0] disk sata:5000cca255dc42d7 /dev/sdl 6.00TB [6:0:0:0] disk sata:5001b44ef905211e /dev/sdm 240GB [8:0:0:0] disk sata:5000cca24dcc471a /dev/sdn 6.00TB [11:0:0:0] disk sas:0x4433221100000000 /dev/sdb 6.00TB [11:0:1:0] disk sas:0x4433221101000000 /dev/sdc 6.00TB [11:0:2:0] disk sas:0x4433221102000000 /dev/sdd 6.00TB [11:0:3:0] disk sas:0x4433221103000000 /dev/sde 6.00TB [11:0:4:0] disk sas:0x4433221105000000 /dev/sdf 6.00TB [11:0:5:0] disk sas:0x4433221107000000 /dev/sdg 6.00TB [N:0:2:1] disk pcie 0x144d:0xa801 /dev/nvme0n1 500GB [N:1:0:1] disk pcie 0x1b4b:0x1093 /dev/nvme1n1 256GB

My Beta2 short test. Running a normal test now. ShortSyncTestReport_2019_08_13_0042.txt

From my experience using my DiskSpeed app, the sd? drive assignment is done in the Port No order on the controller - so sdc is one port after sdb for example. All drives on a given controller are given sd? ID's and then all drives on the next controller and so on until all controllers have been processed. So if the unRAID disk order is done in sd? order, then it should match up barring the controllers initializing in a different order on boot.

["parity"] idx="0" ["disk1"] idx="1" ["disk2"] idx="2" ["disk3"] idx="3" ["disk4"] idx="4" ["disk5"] idx="5" ["disk6"] idx="6" ["disk7"] idx="7" ["disk8"] idx="8" ["disk9"] idx="9" ["parity2"] idx="29" ["cache"] idx="30" ["cache2"] idx="31" ["flash"] idx="54"

How about as neither? I have a NVMe drive mounted by UD for a VM. EDIT: I see the first two are only for unraid drives, but the last is still an option.

@Pauven - You were wondering if having a saturated controller would affect the results if drives were moved around to maximize bandwidth - looks like it did. File on the 9th is with my PCIe 2 SAS controller saturated, file from the 10th is after I moved two drives to the MB controller. utt.zip

H/W path Device Class Description ================================================================ /0/100/1.1/0.1 storage X399 Series Chipset SATA Controller /0/100/1.1/0.2/4/0 scsi11 storage SAS2008 PCI-Express Fusion-MPT SAS-2 [Falcon] /0/100/1.2/0 storage NVMe SSD Controller SM961/PM961 /0/100/8.1/0.2 storage FCH SATA Controller [AHCI mode] /0/117/1.2/0 storage WD Black NVMe SSD /0/117/8.1/0.2 storage FCH SATA Controller [AHCI mode] /0/1 scsi0 storage /0/2 scsi4 storage /0/3 scsi5 storage /0/4 scsi6 storage /0/5 scsi8 storage /0/6 scsi1 storage /0/7 scsi2 storage /0/8 scsi3 storage

Adding the no nvme option excluded the NVMe drives from the output lsscsi -H --no-nvme

Ran 4.0 through on a short test, it reported the following. Probably related to the above issue I posted. It also reported the same controller 3 times at the end. Completed: 0 Hrs 2 Min 34 Sec. ./unraid6x-tunables-tester.sh: line 200: scsiN: 0[scsibus]: syntax error: invalid arithmetic operator (error token is "[scsibus]") ./unraid6x-tunables-tester.sh: line 201: scsiN: 0[driver]: syntax error: invalid arithmetic operator (error token is "[driver]") ./unraid6x-tunables-tester.sh: line 202: scsiN: 0[name]: syntax error: invalid arithmetic operator (error token is "[name]") ./unraid6x-tunables-tester.sh: line 204: ${#scsiN:0[@]}: bad substitution ./unraid6x-tunables-tester.sh: line 209: scsiN: 0[@]: syntax error: invalid arithmetic operator (error token is "[@]") ./unraid6x-tunables-tester.sh: line 200: scsiN: 1[scsibus]: syntax error: invalid arithmetic operator (error token is "[scsibus]") ./unraid6x-tunables-tester.sh: line 201: scsiN: 1[driver]: syntax error: invalid arithmetic operator (error token is "[driver]") ./unraid6x-tunables-tester.sh: line 202: scsiN: 1[name]: syntax error: invalid arithmetic operator (error token is "[name]") ./unraid6x-tunables-tester.sh: line 204: ${#scsiN:1[@]}: bad substitution ./unraid6x-tunables-tester.sh: line 209: scsiN: 1[@]: syntax error: invalid arithmetic operator (error token is "[@]") ShortSyncTestReport_2019_08_09_1101.zip

Two, it looks like. [N:0] /dev/nvme0 Samsung SSD 960 EVO 500GB S3X4NB0K309**** 3B7QCXE7 [N:1] /dev/nvme1 WDC WDS256G1X0C-00ENX0 17501442**** B35900WD (last 4 of SN removed by me)

I tried running version 4.0 against my system and I noticed some errors displaying before the screen cleared and it asked me if I wanted to continue. Querying lsscsi for the SCSI Hosts ./unraid6x-tunables-tester.sh: line 127: declare: `scsiN:0': not a valid identifier ./unraid6x-tunables-tester.sh: line 128: scsiN:0[scsibus]=N:0: command not found ./unraid6x-tunables-tester.sh: line 130: scsiN:0[driver]=/dev/nvme0: No such file or directory ./unraid6x-tunables-tester.sh: line 127: declare: `scsiN:1': not a valid identifier ./unraid6x-tunables-tester.sh: line 128: scsiN:1[scsibus]=N:1: command not found ./unraid6x-tunables-tester.sh: line 130: scsiN:1[driver]=/dev/nvme1: No such file or directory Querying lshw for the SCSI Host Names, please wait (may take several minutes) nas-diagnostics-20190809-0744.zip

I've got my drive cleared off and ready to be moved to a different controller and ready to test. 👍

The threshold mentioned above might be a way to work around this. 162.5MB/s is only 0.06% faster than 162.4MB/s. Maybe if the speed difference is less than 0.2% but utilizes less memory, take the values that result in less memory. 0.2% would match 162.175MB/s and up. 0.5% would allow up to nearly 1MB/sec difference.

Version 2.1 has been pushed Changelog: Controller Information: Display the Partiy/Disk/Cache label in the port listing if the drive is assigned Controller Benchmark: Display unRAID's Parity/Disk/Cache labels in graph Update Highcharts from 6.0.7 to 7.1.2 Display simultaneous max throughput in the Controller Benchmark graph Add "Avg" to Controller Benchmark graph labels to better clarify the value Dynamically configure Controller Benchmark graph's height based on number of drives attached When editing the drive vendor, display the drive info page instead of the home page on submit