9 minutes ago, fargotto48 said:

Problem software (letter disk changes)

That‘s not an issue at all since Unraid doesn‘t assign disks with drive letters, it assigns them based on their UUID.

As said this is most likely a driver issue.

 

On my system the Kernel also reasigns sdX for the disks sometimes when I adr or remove a disk but that‘s not the issue!

 

Please post your Diagnostics when only 3 disks show up (since there will be most likely only three disks be showing up in the system and if thats the case then it‘s a driver issue).

I bet the driver has something to do with that, you have to wait for an answer on GitHub.

2 hours ago, ich777 said:

Please post your Diagnostics when

Wow, I'm finally home.

Here's my data archive.

olips-nas01-diagnostics-20241004-2025_.zip

Here's another one. Immediately after the reboot

hpsa_update_device_info: inquiry failed, device will be skipped hpsahba

olips-nas01-diagnostics-20241004-2044.zip

Edited October 4, 20241 yr by fargotto48

24 minutes ago, fargotto48 said:

Here's my data archive.

As you can see here only three disks are recognized:

 

Here is also something in the logs which indicate a driver issue:

Oct  4 05:36:03 Tower kernel: hpsa can't handle SMP requests
Oct  4 05:36:03 Tower kernel: hpsa 0000:06:00.0: aborted: LUN:000000c000000001 CDB:12000000310000000000000000000000
Oct  4 05:36:03 Tower kernel: hpsa 0000:06:00.0: hpsa_update_device_info: inquiry failed, device will be skipped.
Oct  4 05:36:03 Tower kernel: hpsa 0000:06:00.0: aborted: LUN:000000c000000101 CDB:12000000310000000000000000000000
Oct  4 05:36:03 Tower kernel: hpsa 0000:06:00.0: hpsa_update_device_info: inquiry failed, device will be skipped.
Oct  4 05:36:03 Tower kernel: hpsa 0000:06:00.0: aborted: LUN:000000c000000201 CDB:12000000310000000000000000000000
Oct  4 05:36:03 Tower kernel: hpsa 0000:06:00.0: hpsa_update_device_info: inquiry failed, device will be skipped.
Oct  4 05:36:03 Tower kernel: hpsa 0000:06:00.0: aborted: LUN:000000c000000301 CDB:12000000310000000000000000000000
Oct  4 05:36:03 Tower kernel: hpsa 0000:06:00.0: hpsa_update_device_info: inquiry failed, device will be skipped.
Oct  4 05:36:03 Tower kernel: hpsa 0000:06:00.0: aborted: LUN:000000c000000401 CDB:12000000310000000000000000000000
Oct  4 05:36:03 Tower kernel: hpsa 0000:06:00.0: hpsa_update_device_info: inquiry failed, device will be skipped.
Oct  4 05:36:03 Tower kernel: hpsa 0000:06:00.0: scsi 3:0:0:0: added RAID              HP       P410             controller SSDSmartPathCap- En- Exp=1
Oct  4 05:36:03 Tower kernel: hpsa 0000:06:00.0: scsi 3:0:1:0: masked Direct-Access     ATA      MM1000GFJTE      PHYS DRV SSDSmartPathCap- En- Exp=0
Oct  4 05:36:03 Tower kernel: hpsa 0000:06:00.0: scsi 3:0:2:0: masked Direct-Access     ATA      MM1000GFJTE      PHYS DRV SSDSmartPathCap- En- Exp=0
Oct  4 05:36:03 Tower kernel: hpsa 0000:06:00.0: scsi 3:0:3:0: masked Direct-Access     ATA      MM1000GFJTE      PHYS DRV SSDSmartPathCap- En- Exp=0
Oct  4 05:36:03 Tower kernel: hpsa 0000:06:00.0: scsi 3:0:4:0: masked Enclosure         PMCSIERA  SRC 8x6G        enclosure SSDSmartPathCap- En- Exp=0
Oct  4 05:36:03 Tower kernel: hpsa can't handle SMP requests

 

 

As I thought this is not a software issue this is a driver issue.

5 minutes ago, ich777 said:

As I thought this is not a software issue this is a driver issue.

So, the plugin itself is buggy?


 

And this is NOT a problem with the disks.

It feels like the plugin only sees 3 disks bay out of 8.

I didn't find out which ones exactly. But it's definitely not a problem with the disks, since other disks will be visible if I swap them (3 pieces)

Edited October 4, 20241 yr by fargotto48
disk

20 minutes ago, fargotto48 said:

So, the plugin itself is buggy?

Maybe, but again, I can‘t help with that since my plugin is only installing the precompiled driver for Unraid for the correct Kernel version and does nothing more.

 

You should definitely post the output from above on the GitHub issue that you‘ve created.

 

Maybe the firmware from your controller is outdated or something like that but these are only assumptions since I really can’t tell what‘s going on, again, I have no hardware to test and I did this plugin on request.

 

You should really ask your questions on GitHub.

7 minutes ago, ich777 said:

You should really ask your questions on GitHub.

You're a cool dude) Maybe in the future it will be possible to buy you this garbage from ebay for $ 10?))

for any tests

Edited October 4, 20241 yr by fargotto48

@fargotto48 you should also look at this thread (a simple Google search have brought that up, the tread is in German hut I think you can translate it) :

 

Please also note that he linked a Youtube video and he also said that his firmware on the controller was outdated.

8 minutes ago, fargotto48 said:

Maybe in the future it will be possible to buy you this garbage from ebay for $ 10?))

No thank you… 😂

I‘m happy with my ASM1166 controller… I really try to avoid such problematic hardware…

 

I really don‘t need such a controller and AFAIK these are usually only found in HP servers.

16 minutes ago, ich777 said:

and AFAIK these are usually only found in HP servers.

I downloaded the proprietary utility HP Smart Array Assistant. But the problem starts EARLIER than the system loads.

That is, the program itself "sees" only 3 disks (bed 6,7,8). In fact, it only sees beds 6,7,8.

And my firmware is the latest. I don't know, I'll try to do a clean slate according to the guide

That is, the problem is much deeper. After rebooting, most likely the controller itself sees that "something is wrong". And goes crazy.

After all, the HBA mode is not actually officially supported on the p410 and p410i

 

On Monday, when I come to work, I will try to remove the 1GB cache board and the super capacitor.

Edited October 5, 20241 yr by fargotto48

On 10/4/2024 at 9:35 PM, ich777 said:

I really don‘t need such

In general, I think so, after rebooting, the p410 controller, although it sees that it is in HBA mode, "doesn't like" it. And it "breaks" or puts some of the disks into emergency mode.

Conventionally, this problem can be solved by DISABLE the boot initialization of the p410. So that it does not load at all, but is seen as a "dumb" adapter in the system.

There is no such option on the old HP Gen5 server, I cannot disable the PCIe devices.

Therefore, I simply made 3 RAID10 mirrors again.

Pros: the battery and 1 GB cache work, the disks work together (in UNRAID, all disks work on their own and the speed is not summed up).

But there is a MINUS - I DO NOT KNOW whether the disk is out of order, because of the logical array. I need to go to the server room and look or reboot remotely and look.

But I don't think it's a problem, it's not my main storage, it's mostly a trash heap

And in principle, p410 does NOT officially support this. It's just a "hack" and it is not required to work correctly. I'm not particularly upset.

The question is different. Can I disable the warning about the absence of Parity? (I don't need it, my RAID controller does this).

As an array, I have a RAID10 300 GB mirror made of two SAS disks.

I will throw the other 2 RAID10 mirrors on a remote computer via iSCSI.


I would like to remove Parity so that the green light is on instead of the "triangle"

16 minutes ago, fargotto48 said:

As an array, I have a RAID10 300 GB mirror made of two SAS disks.

This is kot the recommended way of doing it on Unraid and out of scope for this thread.

 

16 minutes ago, fargotto48 said:

I would like to remove Parity so that the green light is on instead of the "triangle"

This is not possible and not how Unraid works, this is also out of scope for this thread.

 

10 minutes ago, fargotto48 said:

Or just make a "hack" so that the green light is on for me. Otherwise, the triangle is an eyesore.

Again, out of scope for this thread.

 

Please also note that you can edit posts a forum is not a chat, you completely bloat this thread with unecessary messages ehich could be easily combined into one.

You now have almost two pages in this thread which are just messages from you.

6 minutes ago, ich777 said:

Также обратите внимание, что вы можете редактировать сообщения, поскольку форум

I understand that I'm trying to "stretch an owl onto a globe".

And the meaning of UNRAID is different (flexible disk replacement, own raid logic).

Sorry for the off-topic

 

Unraid is simply too cool a system (virtualization, docker, applications, frequent updates).

Even for non-standard solutions, it is better than OpenMediaValute (complete crap), or TrueNAS (yes, slightly better, but it is not the same)

Edited October 5, 20241 yr by fargotto48

Has anyone successfully had the Hailo drivers plugin working on UnRAID? I need to reboot to get diagnostics etc as I removed it to get the server booted, but System Devices showed the Hailo CoProcessor in devices, but the driver reported "No Device Found" and the Main screen doesn't load (Assuming the Hailo homescreen widget is failing).

 

I will reinstall it over the weekend and generate diagnostics etc, just wondered if anyone had it working OK?

Edited October 9, 20241 yr by Flubster
spelling

17 minutes ago, Flubster said:

Has anyone successfully had the Hailo drivers plugin working on UnRAID? I need to reboot to get diagnostics etc as I removed it to get the server booted

I don‘t understand what you mean with that, was your server not booting with the Hailo plugin installed or was the server crashing?

 

My test server boots fine with the Hailo 8L last time I checked.

 

Maybe the executable is broken that detects the Hailo devices. I‘ll look i to that.

 

On what Unraid version are you? Can you please post a bit more details about your server or maybe include Diagnostics?

server boots, but the dash fails to load. Just a white page after login, which I've assumed is the widget failing. Digging in a little I see its doing "hailostatus -json" therefore i'll get this via terminal when its booted. I did manage to get into the plugin tab via direct URL, which showed no device found. Will send diagnostics when I plug the device back in.

 

6.12.13 btw.

2 minutes ago, Flubster said:

server boots, but the dash fails to load. Just a white page after login, which I've assumed is the widget failing. Digging in a little I see its doing "hailostatus -json" therefore i'll get this via terminal when its booted. I did manage to get into the plugin tab via direct URL, which showed no device found. Will send diagnostics when I plug the device back in.

What do you mean dash fails to load? The entire webui or only the plugin page?

Are you sure that this is not actually a browser issue

 

I think somehow the application is broken that reports the device stats, maybe I remove it for now.

Just do:

ls -la /dev/hailo*

from a Unraid terminal, if you get a output with something like /dev/hailo0 then your device is properly recognized and working.

 

If you get nothing back then your device is not working, granted you've installed the driver and already rebooted.

Hi,

 

Checked and not browser issue, the "Dashboard" tab does not load. Other tabs do, just not dashboard. 

 

Linux 6.1.106-Unraid.
root@abyss:~# ls -la /dev/hailo*
crw------- 1 root root 241, 0 Oct  9 17:46 /dev/hailo0

 

Suggesting it is loading? curiously am getting

 

[ 1611.268879] hailo 0000:03:00.0: Invalid general ioctl code 0x400c6705 (nr: 5)

 

spamming every 30 seconds.

 

[   39.299708] hailo_pci: loading out-of-tree module taints kernel.
[   39.301608] hailo: Init module. driver version 4.18.0
[   39.302268] hailo 0000:03:00.0: Probing on: 1e60:2864...
[   39.302932] hailo 0000:03:00.0: Probing: Allocate memory for device extension, 11632
[   39.303510] hailo 0000:03:00.0: enabling device (0000 -> 0002)
[   39.304205] hailo 0000:03:00.0: Probing: Device enabled
[   39.304839] hailo 0000:03:00.0: Probing: mapped bar 0 - 000000005f57c0bd 16384
[   39.305337] hailo 0000:03:00.0: Probing: mapped bar 2 - 00000000b5b99e8b 4096
[   39.305696] hailo 0000:03:00.0: Probing: mapped bar 4 - 0000000001bd3bb8 16384
[   39.305962] hailo 0000:03:00.0: Probing: Setting max_desc_page_size to 4096, (page_size=4096)
[   39.306223] hailo 0000:03:00.0: Probing: Enabled 64 bit dma
[   39.306534] hailo 0000:03:00.0: Probing: Using userspace allocated vdma buffers
[   39.306783] hailo 0000:03:00.0: Disabling ASPM L0s
[   39.307083] hailo 0000:03:00.0: can't disable ASPM; OS doesn't have ASPM control
[   39.307496] hailo 0000:03:00.0: Successfully disabled ASPM L0s
[   39.456253] hailo 0000:03:00.0: Firmware was loaded successfully
[   39.469431] hailo 0000:03:00.0: Probing: Added board 1e60-2864, /dev/hailo0
[  488.950637] hailo 0000:03:00.0: Invalid general ioctl code 0x400c6705 (nr: 5)

 

Diagnostics attached.

abyss-diagnostics-20241009-1822.zip

OK, problem is the 4.18.0 driver, if I revert to 4.17.1 it is working correctly.

On 10/9/2024 at 7:41 PM, Flubster said:

OK, problem is the 4.18.0 driver, if I revert to 4.17.1 it is working correctly.

I have a similar problem that my dashboard is not loading. Can you explain detailed how to revert den driver if its the same issue on my unraid server?

Go here

 

http://<unraid-server-here>/Settings/hailort-driver

 

Change driver version. Reboot.

 

Update driver as ich777 has removed the dependencies and dashboard widget recently.

 

4.18.0 doesn't work at all. I notice 4.19.0 is out now so suspect that will compiled soon.

Edited October 26, 20241 yr by Flubster
Spelling

Okay I don't have this driver installed. So it might be something different on my end. Will try to check the logs if there is something written. All worked before I did updated my plugins including this one

3 hours ago, Flubster said:

4.18.0 doesn't work at all.

This is not true, the version available for Unraid does work indeed but the software which was responsible for grabbing the data wasn‘t working properly.

 

4.19.0 is already available for Unraid 7.0.0-beta.4