IG-88

Driver extension jun 1.03b/1.04b for DSM6.2.2 for 3615xs / 3617xs / 918+

Recommended Posts

2 minutes ago, IG-88 said:

you forgot to mention what type or loader you use

i guess its 918+ and the intel will work for sure with the new 0.7+, just wait

My fault, it's indeed 918+ with 1.04b. I'm getting excited 😄
Thank you in advance for all the work !

Share this post


Link to post
Share on other sites
On 1/9/2020 at 12:19 PM, IG-88 said:

dvb drivers are not part of my package

please open a new thread for this kind of problem

 

you should ask the person who made this modules (presumably for dsm 6.2.0) if he is willing to redo the drivers for 6.2.2

i've already documented the change thats needs to be done when compiling new drivers for 6.2.2

https://xpenology.com/forum/topic/7187-how-to-build-and-inject-missing-drivers-in-jun-loader-102a/?do=findComment&comment=122631

 


Sorry, I already know they are not network or storage drivers. But I asked you because I think you are the most knowledgeable and indicated.

No person in this forum made these modules. I made them myself.
And these modules work well in DSM 6.2.0, but in DSM 6.2.2 the last module does not work.

I have read your message from the other thread:
 

On 10/2/2019 at 6:27 PM, IG-88 said:

synology changed the default kernel config and as they don't publish any source we can only guess and try

the functions in question are part of PCIE_ASPM so you will have to disable CONFIG_PCIEASPM in the kernel config when compiling modules for dsm 6.2.2, tested it and my freshly compile modules for 918+ are loaded and work (i'm about to continue the extra.lzma's in the next day's starting with 918+)

ifair jun wrote something here

https://xpenology.com/forum/topic/13074-dsm-621-23824-warning/?tab=comments#comment-95497


 

Edit: for the old 918+ beta 4.4.x kernel the Makefile needs to be modified

VERSION = 4
PATCHLEVEL = 4
SUBLEVEL = 59
EXTRAVERSION =+

 

the chroot enviroment should also get additional packages

 apt-get install openssl libssl-dev

 

for 6.2.2 i also added signature files

generate certs like here: https://wiki.gentoo.org/wiki/Signed_kernel_module_support

place keyfiles in kerne root, for kernel 4.4.x in /certs instead of kernel root

copy one file to a new name: cp signing_key.x509 signing_key

create empty files in kernel root: extra_certificates, extra_certificates_untrusted, trusted_certificates, untrusted_certificates

 


I have no idea of doing that process, I have searched Google and I find little information.

Exactly where can I disable CONFIG_PCIEASPM in the kernel config? (Is that a system file?)
I compile the modules in DSM 6.2.2, in Debian Chroot. Is the kernel config a file inside Debian Chroot?

And you know if the signature files/certs are necessary or optional? (because the first time you mention that you have only modified the kernel config, and it worked for you)

By the way, I use DS3615xs, not 918+.

Thank you again for answering me and for all your help! 👍
Regards!

Edited by manuelin

Share this post


Link to post
Share on other sites
14 minutes ago, manuelin said:

Exactly where can I disable CONFIG_PCIEASPM in the kernel config? (Is that a system file?)
I compile the modules in DSM 6.2.2, in Debian Chroot. Is the kernel config a file inside Debian Chroot?

And you know if the signature files/certs are necessary or optional? (because the first time you mention that you have only modified the kernel config, and it worked for you)

 

please put your questions in the section where you took the content from, discussing this here does not help (off topic)

people interested in compiling drivers might be interested too or might profit from that later when its in the place where it belongs to and has the right context

 

  • Like 1

Share this post


Link to post
Share on other sites
9 minutes ago, IG-88 said:

 

please put your questions in the section where you took the content from, discussing this here does not help (off topic)

people interested in compiling drivers might be interested too or might profit from that later when its in the place where it belongs to and has the right context

 

 

Okay, you're right, sorry!
Now I quote your message in the other thread.

Thanks again ;)

Share this post


Link to post
Share on other sites
4 hours ago, lemon55 said:

Me too usb ups is not working.
6.1.7 works is fine, ups is APC BE700G-RS
Where can I see the logs to send you?

 

tested 3615, 6.2.2 u4, last extra.lzma from mid. december

apc back-ups 800 - after enabling ups support in dsm the deviceinfo button shows the information about the ups

so its not a general problem, maybe a usb problem?

try a different usb port

logs (dmesg?) can go as max. 10MB here or you can upload them somewhere and use a link

 

Share this post


Link to post
Share on other sites

Hi All,

 

Im trying to get a HPE ProLiant DL360e Gen8 (got a couple of these machines for free and want to use them for my lab) to work with the XPenology project, the server contains following.

 

2 * Intel(R) Xeon(R) CPU E5-2450L 0 @ 1.80GHz  (Sandy Bridge)

1 * Intel(R) i350 Gigabit 4-port Network Adapter

1 * HPE NC523SFP 10Gb 2-port Server Adapter (QLogic 82XX)

Intel's C600 chipset

HPE Smart Array P420 Controller

 

It boots up, but can not find any disks (there are 8 * 2tb in HBA mode attached), and the extra.lzma for DS3617xs has been added to the usb stick.

 

Are there any way to login to the device during install ? and check what drivers are loaded ?

 

i want to check if the hpsa.ko is loaded, or perhaps it dosnt recognize the controller.

 

Kind Regards,

lthasenhod

Share this post


Link to post
Share on other sites

Just a quick question :  I bought a 6 tb seagate SAS drive .  Will this work with this version of DS3617 1.03b  ???

 

Edited by painkiller895

Share this post


Link to post
Share on other sites

 

On 1/16/2020 at 4:40 PM, lthasenhod said:

2 * Intel(R) Xeon(R) CPU E5-2450L 0 @ 1.80GHz  (Sandy Bridge)

as of the number of cores 3617 should be used

 

On 1/16/2020 at 4:40 PM, lthasenhod said:

1 * Intel(R) i350 Gigabit 4-port Network Adapter

should be no problem with the latest extra.lzma from december

 

On 1/16/2020 at 4:40 PM, lthasenhod said:

1 * HPE NC523SFP 10Gb 2-port Server Adapter (QLogic 82XX)

 

driver qlcnic (QLOGIC QLCNIC 1/10Gb Converged Ethernet)

driver present, not sure if i ever saw feedback about this driver and you hp nic, in theory it should work as we have a driver and if it does not we can try to compile one from external source

 

On 1/16/2020 at 4:40 PM, lthasenhod said:

Intel's C600 chipset

 

driver isci.ko, as sas driver present but not tested, the chipset as for the cpu does not need drivers to work

 

On 1/16/2020 at 4:40 PM, lthasenhod said:

HPE Smart Array P420 Controller

 

in general a problem as external source drivers never worked, afair 410 or 420 i can be set into it mode (later 4xx might have it in controller bios older might need to be switched with a tool)

in normal raid mode it might not work (see comment about scsi/sas drivers in 1st post), imho pointless in raid mode, never saw anyone even capable of trying in it mode (hba mode)

looks like the p420 needs to be switched to hba mode

https://communities.vmware.com/thread/484114

hpsa.kodriver is present if someone want to test

 

as a simple test you can define one drive as sigle disk raid and see if it show up, if that works you can try to bring the controller to it/hba mode und see if disks will be found (without defining a raid)

 

i'm not going to try anything with drivers as long as i dont own a p420 and atm they are not cheap on amazon and we dont have any of those at work (at least not as pcie card)

 

 

3 hours ago, painkiller895 said:

Just a quick question :  I bought a 6 tb seagate SAS drive .  Will this work with this version of DS3617 1.03b  ???

 

never tried one myself but there are people using hpe proliants and dell server and those often come with sas drives

 

Share this post


Link to post
Share on other sites
On 1/11/2020 at 4:19 PM, Setchen said:

 

Hi Nadiva,

 

could you, please, update the DSM Post-Installation topic with your successful achievement?

Also, please, let us know, please, how you manage to add a NVMe module on the Gen10 AMD microserver? With an PCI x1, x4, x8 card?

Thanks a lot!

Hello to the owners of the best NAS server ever. The one and only AMD 12watt 5000passmark device, more than twice as powerful as DS918+ while quieter and lower on energy and at the same price.

After installing the great IG-88's image, i just activated NVME with libNVMEpatch and added 10gbit with synoinfo NIC patch. This is my first NAS install (after ovm, freenas, proxmox etc) which didn't crash or didn't need a single restart since set up! No matter what was configured. To my shock, DSM manages its quietness even during transcoding and busy times. All other OS made it shout loud in an annoying sinusoid sound profile.

I have a basic Intel 1800MB/s NVME (cheapier than sluggish SSDs of the same size) on a PCIx8, on a generic PCI to NVME adapter and cheap 10gbit Aquantia on PCIx1 (it is PCIx4 but only 1 active lane). Since it's PCI3 it still gives us power thirsty maniacs beautiful 1000MB/s (still i noticed Gen8 owners in other forums moan they will never buy Gen10, stuck in 1gbit era, they should try). Rationally we give the better bus to NVME because it can get benefit of faster internal transfers, while 10gbit will likely won't reach top speed most of the day. There's just a little loss of bandwith. I've put a passive cooler plate on it, it boasts incredible low 24C temperature when idle. So definitely both PCIs can be nicely utilized, 1x is not for low priority devices at all, i attached sound card via USB in order to have notifications loud in future.

But i wanted more. I wasn't happy with NVME cache limit of 440GB and possible cache rebuild after reboot (which didn't happen) and possible leak of unencrypted data (not sure if it's the case, can't find out if @ecryptfs@ files are copied there or not) so I've set up a NVME volume:

Quote


synopartition --part /dev/nvme0n1 12

mdadm --create /dev/md4 --level=1 --raid-devices=1 --force /dev/nvme0n1p3

mkfs.btrfs -f /dev/md4

 

And here's the 2TB NVME working hard to deliver hot data (user data except movies), and being synced to RAID (cold data) in order to create another layer of redundancy like in lazy RAID systems (Unraid, SnapRaid) (attachment). 😋 That was my dream setup like I did replication in ZFS, which however crashed the drives completely and the OS too. In that moment, I looked into IG-88 new messages while wiping the old OS, and realized we can now have a maxed out DSM setup.

The last thing to solve is to make RAID drives idle. Probably need to dig deep to agents which maintain the annoying 2GB RAID on each drive. And full disk encryption, but no package in alt repositories.

NVME as drive.png

Share this post


Link to post
Share on other sites
9 hours ago, nadiva said:

After installing the great IG-88's image,

 

<slight cough> i'm just offering additional drivers, no loader, no hacking ;-)

 

9 hours ago, nadiva said:

Hello to the owners of the best NAS server ever.

 

https://www.synology.com/en-global/company ?

 

9 hours ago, nadiva said:

added 10gbit with synoinfo NIC patch.

 

thats not really permanent, it will not survive a major update with a full dsm version (200-300MB *.pat file) - same es changing the max disk's manually that way

but being back to 2 nic's after a update has not that much consequences as "reducing" the number of disks and breaking the raid

to find the 2 working nic's you only need to plug the network cable one after another and use ping to find the two still working

 

9 hours ago, nadiva said:

To my shock, DSM manages its quietness even during transcoding and busy times. All other OS made it shout loud in an annoying sinusoid sound profile.

 

and the reason is not a better management of fan's its the lack of it as dsm only knows the of synologys own hardware and has no drivers for most of the silicon reading temp. and managing fan speed

some people with older hp microserver see the different from what you have, fans at max. speed because of the lack of management (hp usually maxes cooling in bios until there is proof there is management, like reading temp. and reducing fan speed - less possible hardware damage if things go wrong)

 

10 hours ago, nadiva said:

i attached sound card via USB in order to have notifications loud in future.

 

dsm can use this? would be news to me

did you test that?

 

10 hours ago, nadiva said:

The last thing to solve is to make RAID drives idle. Probably need to dig deep to agents which maintain the annoying 2GB RAID on each drive. And full disk encryption, but no package in alt repositories.

 

maybe there is something in the tools already there, synology often seems to go there own ways. so looking into the syno scripts and tools might show something

there are all flash units from synology and you can open the *.pat files (and hda1) with 7zip so you can have a look in this too, if there is anything then its in there

 and with some luck the stuff is already in your system too as synology mostly uses one source to build all systems and using switches to switch on/off features (like in synoinfo.conf)

Share this post


Link to post
Share on other sites

Will the new 0.7 work with the j5005 with the newest dsm version, transcoding and network? And if you need a tester for that platform, i can help

Share this post


Link to post
Share on other sites
9 hours ago, IG-88 said:

 

<slight cough> i'm just offering additional drivers, no loader, no hacking ;-)

 

 

https://www.synology.com/en-global/company ?

 

 

thats not really permanent, it will not survive a major update with a full dsm version (200-300MB *.pat file) - same es changing the max disk's manually that way

but being back to 2 nic's after a update has not that much consequences as "reducing" the number of disks and breaking the raid

to find the 2 working nic's you only need to plug the network cable one after another and use ping to find the two still working

 

 

and the reason is not a better management of fan's its the lack of it as dsm only knows the of synologys own hardware and has no drivers for most of the silicon reading temp. and managing fan speed

some people with older hp microserver see the different from what you have, fans at max. speed because of the lack of management (hp usually maxes cooling in bios until there is proof there is management, like reading temp. and reducing fan speed - less possible hardware damage if things go wrong)

 

 

dsm can use this? would be news to me

did you test that?

 

 

maybe there is something in the tools already there, synology often seems to go there own ways. so looking into the syno scripts and tools might show something

there are all flash units from synology and you can open the *.pat files (and hda1) with 7zip so you can have a look in this too, if there is anything then its in there

 and with some luck the stuff is already in your system too as synology mostly uses one source to build all systems and using switches to switch on/off features (like in synoinfo.conf)

yep, thanks for that.

heh :) i might buy a syno for family, but for me  i wouldn't call their retail machines great, as they throw outlet CPUs inside, no ECC, no full disk encryption, security is poor out of the box, running services under root, forcing Plex to be an user sharing the data with real users (who are forced to be part of users group), etc etc. but but all of this can be tweaked in a persistent way.

same with the patches - i find it trivial to make it persistent even after upgrades. even other OS deleted some of my tweaks, actually exactly the same - e.g 10gbit NIC, so i scheduled it using their persistent scheduler. regarding NVME, hearing people have no problem running after minor and major upgrades!

that's why i love DSM on custom machine. replace rubbish hardware with decent one, harden the security, and enjoy the greatness.

re noise, i'm lucky with it, but it seems DSM is throttling the speed down as well. all seems to be fast generally, but e.g. openssl speed is not fast as it should be. no impact of fan speed mode in Hardware & Power panel. also no luck with synofandebug, synofanconfig, synotune. oh wait, why do i have memory compression on!

re audio, yes it works, plug in some 5$ usb sound card with $2 speakers attached, and AudioStation will play music from your custom machine asap. to test, use:

/volume1/@appstore/AudioStation/bin/mplayer <file.mp3>

and now that we have this command, we can load a notification package from one of the two know alt repositories, and ta da... you have custom notifications with sound. 🔊

i'm customizing what i can: plex trakt  total sync for free, ssh via certificate even with homedrives unlocked, mandatory parity check postpone, shell with autocomplete/autocorrect/syntaxhighlightning, update certificate automatically into all apps and make it valid for external domain and on LAN at the same time without ever confirming in browser! etc. if you're interested, let me know.

 

Edited by nadiva

Share this post


Link to post
Share on other sites
43 minutes ago, nadiva said:

and now that we have this command, we can load a notification package from one of the two know alt repositories, and ta da... you have custom notifications with sound. 🔊

 

nice, i did not know about that, whats the package

Share this post


Link to post
Share on other sites
20 minutes ago, IG-88 said:

 

nice, i did not know about that, whats the package

so you can just use synodsmnotify @users title message, or enhance use with community packages "Growler", "Notification Messenger", "Notification forwared", "Notification Essentials". And add own sound to it. Ideally, i'd like a package which can read the message loud:) Anyways, here's the notification with sound (pic):

000395 2020-01-18.png

Share this post


Link to post
Share on other sites
4 hours ago, nadiva said:

re noise, i'm lucky with it, but it seems DSM is throttling the speed down as well. all seems to be fast generally, but e.g. openssl speed is not fast as it should be. no impact of fan speed mode in Hardware & Power panel. also no luck with synofandebug, synofanconfig, synotune. oh wait, why do i have memory compression on!
 

Gen10 finding. See my story here. Indeed after boot, performance is lower and fans never scream. I had to tweak it  that way (now added as post boot script) in order to get the following improvement:

  • openssl speed -elapsed -evp aes-256-xts     from 1200000 to 3200000
  • OPENSSL_ia32cap="~0x200000200000000" openssl speed -elapsed -evp aes-256-gcm       from 60000 to 160000
  • hdparm -Tt /dev/nvme0n1     from 1200 to 2500MBs
  • dd if=/volume3/moviefile of=/dev/zero bs=4G count=1       from 600 to 1200MB/s

same idle power consumption, but 3x faster calcs, 2x faster drives and some fan noise under load. If anyone has openssl xts last number 1mil instead of 3, you better check this.

seems acpi-cpufreq driver, the only one available as seein in ls /usr/lib/modules/*cpu*, is confused here, and sets low TPU max (12W), but the CPU can boost to 35W when needed. Other cpu freq drivers (not governers) for AMD are powernow-k8 and pcc-cpufreq, the first one supposed to by replaced by acpi-cpufreq, the later one is "recomended for ProLiant servers". So i looked into my collection of DSM test files, but there isn't one.

this also gives us idea how full disk encryption could be fast - up to 3.2GBs! Not far from i7 Pro desktop performance. I personally have disk encryption as fast as openssl benchmark (xts or core count * cbc/gcm bench) on any device incl router. (My router provided me encrypted 1gbit shares for years at full speed, that's why i find it a complete joke to buy a NAS which can do 1gbit encyption only.) here you have evidence how poor are all file encryptions.

for those who dont believe how poor is file encryption, see the numbers:

dd if=/dev/zero of=/volume3/normalsharedfolder/test bs=4G count=1                                                                             
0+1 records in
0+1 records out
2147479552 bytes (2.1 GB) copied, 1.90531 s, 1.1 GB/s
dd if=/dev/zero of=/volume3/encryptedsharedfolder/test bs=4G count=1                                                                         
0+1 records in
0+1 records out
2147479552 bytes (2.1 GB) copied, 1.88765 s, 184 MB/s

cached file copy same. samba surprisingly faster - 300MB/s. puzzling as no physical writes happen when copy done, no cache effect. still encryptfs is not good. Full disk encryption would do 1.0-1.1GB/s. No doubt Qnap doesn't use this. In the era of native encryption coming to ZFS, it's shocking Synology would use this obsolete method.

Share this post


Link to post
Share on other sites
On 1/18/2020 at 2:27 AM, nadiva said:

To my shock, DSM manages its quietness even during transcoding and busy times.

 

Are you saying that the Gen10 AMD/APU (Radeon Pro WX 2100) is detected with the 918+ loader, and that you can transcode video?

 

On 1/18/2020 at 2:27 AM, nadiva said:

---> I have a basic Intel 1800MB/s NVME (cheapier than sluggish SSDs of the same size) on a PCIx8, on a generic PCI to NVME adapter...

 

---> But i wanted more. I wasn't happy with NVME cache limit of 440GB and possible cache rebuild after reboot (which didn't happen) and possible leak of unencrypted data (not sure if it's the case, can't find out if @ecryptfs@ files are copied there or not) so I've set up a NVME volume:

And here's the 2TB NVME working hard to deliver hot data (user data except movies), and being synced to RAID (cold data) in order to create another layer of redundancy like in lazy RAID systems (Unraid, SnapRaid) (attachment). 😋 That was my dream setup like I did replication in ZFS, which however crashed the drives completely and the OS too. In that moment, I looked into IG-88 new messages while wiping the old OS, and realized we can now have a maxed out DSM setup.

 

NVME as drive.png

 

So you've transformed the 2To NVMe from a read-only cache to a SSD disk or to a read/write SSD? due to cache limit?

What's the configuration to set up in DSM to sync the hot and cool data?

Share this post


Link to post
Share on other sites

Hi, thanks for the guide, I have just installed sucessfully a microserver gen10 using instructions for loader ds3615xs, without an intel network card, ie its using internal networking.  I will now test same on microserver gen7 and microserver gen8 and will post results

Edited by djvas335

Share this post


Link to post
Share on other sites
1 hour ago, Setchen said:

 

Are you saying that the Gen10 AMD/APU (Radeon Pro WX 2100) is detected with the 918+ loader, and that you can transcode video?

 

 

So you've transformed the 2To NVMe from a read-only cache to a SSD disk or to a read/write SSD? due to cache limit?

What's the configuration to set up in DSM to sync the hot and cool data?

No, i ment that those are the busiest times to produce noise. I don't think my GPU was ever used and honestly i don't care:

Spoiler

 

- GPU is Radeon R7 800MHz @ 1000 gpu passmarks

- GPU you mention WX2100 is an extension you must buy @ 2200 gpu passmarks

- both performances are poor compared to my laptop GT1060 @ 7000 or Titan RTX @ 16000

- it's not clear if any of them can be actually used in any OS as a transcoder

- it's completely undetected in DSM as of now (lspci, dmidecode, dmesg)

- i suspect CPU alone is faster then GPU for transcoding

- it's ment for 4K displaying, e.g. connect a big display and show your cameras there permanently

- i'm happy this GPU provides me displayport so i can easily switch to it with KVM hub or put in as picture-in-picture on main screen

- my "transcoding" is plex analyzing, i.e. creating thumbnails offline and the (cpu) speed is very good (~100movies an hour)

- realtime "transcoding" is something that should not exist imo, i can't see any stationary/mobile device which needs that e.g. mobile plays 4K HDR DTS HEVC easily via 300+mbps wifi
- when i checked the performance of realtime cpu transcoding, which Plex somehow loves to do sometimes, it was ok for HD movies. but for 4K HDR even 100W card wouldn't stand a chance.. so why bother

 

 

Yes I actually never tried the cache and just formatted it into a volume due to

- cache limit, not utilizing most of the space

- i love main data to reside on high IOPS high sequential speed NVME drive, then replicate to RAID, that's why i bought it, not for a cache which canshould be substituted with RAM

- i didn't move all stuff yet but later i want incremental backup with snapshots, ZFS/BTRFS style send->receive, and this is done with 'Snapshot replication' foremost or 'HyperBackup'. All other backups are amateurish, they don't send changed blocks (in ZFS even encrypted), and snapshots for double recovery from each drive, and even system files.  This is where the beauty of ZFS/BTRFS is

Share this post


Link to post
Share on other sites
17 minutes ago, nadiva said:

No, i ment that those are the busiest times to produce noise. I don't think my GPU was ever used and honestly i don't care:

  more blabs (Reveal hidden contents)

 

- GPU is Radeon R7 800MHz @ 1000 gpu passmarks

- GPU you mention WX2100 is an extension you must buy @ 2200 gpu passmarks

- both performances are poor compared to my laptop GT1060 @ 7000 or Titan RTX @ 16000

- it's not clear if any of them can be actually used in any OS as a transcoder

- it's completely undetected in DSM as of now (lspci, dmidecode, dmesg)

- i suspect CPU alone is faster then GPU for transcoding

- it's ment for 4K displaying, e.g. connect a big display and show your cameras there permanently

- i'm happy this GPU provides me displayport so i can easily switch to it with KVM hub or put in as picture-in-picture on main screen

- my "transcoding" is plex analyzing, i.e. creating thumbnails offline and the (cpu) speed is very good (~100movies an hour)

- realtime "transcoding" is something that should not exist imo, i can't see any stationary/mobile device which needs that e.g. mobile plays 4K HDR DTS HEVC easily via 300+mbps wifi
- when i checked the performance of realtime cpu transcoding, which Plex somehow loves to do sometimes, it was ok for HD movies. but for 4K HDR even 100W card wouldn't stand a chance.. so why bother

 

 

Yes I actually never tried the cache and just formatted it into a volume due to

- cache limit, not utilizing most of the space

- i love main data to reside on high IOPS high sequential speed NVME drive, then replicate to RAID, that's why i bought it, not for a cache which canshould be substituted with RAM

- i didn't move all stuff yet but later i want incremental backup with snapshots, ZFS/BTRFS style send->receive, and this is done with 'Snapshot replication' foremost or 'HyperBackup'. All other backups are amateurish, they don't send changed blocks (in ZFS even encrypted), and snapshots for double recovery from each drive, and even system files.  This is where the beauty of ZFS/BTRFS is

 

Thank Dude!

I appreciate your replies and explanations and would like to know more about the "beauty of ZFS/BTRFS" stuff!

Seems like you know very lots about it! True?

Thanks also for giving tips! You're a great support.

Take care!

 

Also, do you think that installing PHPVirtualbox withnin DSM 918+ on Gen10 baremetal could help benefiting from the Radeon R7 800MHz @ 1000 gpu passmarks?

Edited by Setchen

Share this post


Link to post
Share on other sites

beauty of ZFS/BTRFS beside copy-on-write, checksumming, is what most people don't use even when given for free:

- snapshots which is comfortably visible as Previous Versions in windows Explorer properties

- files not overwritten when replaced, deleted, instead kept as transaction or put to samba thrashbin (+snapshot +replication +Drive versioning +online backup)

forums:
 

Spoiler

 

- i lost my dockers/jails
- sad, do you have snapshots?
- no
- (pause)

 

- efficient operation with blocks, deduplication, instead of rsyncing files on high level, you can copy whole shares as it was a file, in ZFS without reencrypting

- you can create a block device inside of a pool, the VM will think it's Ext4, while you add benefits of your fs on top of this primitive fs

- compression with negligible impact which works on data partitions, i had 30%+ savings

- (ZFSonLinux) native encryption, which i've tested as super fast on debian

 

list of good resources:

xpenology

list of syno ports: https://www.synology.com/en-global/knowledgebase/DSM/tutorial/Network/What_network_ports_are_used_by_Synology_services

list of syno commands: https://github.com/wuseman/SYNOLOGY

my notes full of commands to push limits and break rules of DSM, freenas, OMV :)

 

benchmarks notes (pic):

performance.png

Share this post


Link to post
Share on other sites

Tested virtio module on DS3617 PVE VM,  virtio_net is OK, but virtio_scsi crashed during loading process. Here  is the log captured through virtual serial port:

 

:: Loading module isci ... [  OK  ]
:: Loading module hpsa ... [  OK  ]
:: Loading module nvme ... [  OK  ]
:: Loading module virtio ... [  OK  ]
:: Loading module virtio_ring ... [  OK  ]
:: Loading module virtio_mmio ... [  OK  ]
:: Loading module virtio_pci ... [  OK  ]
:: Loading module virtio_blk ... [  OK  ]
:: Loading module virtio_scsi[    2.269040] BUG: unable to handle kernel NULL pointer dereference at 00000000000001d8
[    2.270007] IP: [<ffffffffa02feb80>] virtscsi_queuecommand_single+0x10/0x30 [virtio_scsi]
[    2.270007] PGD 13684d067 PUD 13680c067 PMD 0 
[    2.270007] Oops: 0000 [#1] SMP 
[    2.270007] Modules linked in: virtio_scsi(F+) virtio_blk(F) virtio_pci(F) virtio_mmio(F) virtio_ring(F) virtio(F) nvme(F) hpsa(F) isci(F) arcmsr(F) mvsas(F) mvumi(F) 3w_xxxx(F) 3w_sas(F) 3w_9xxx(F) aic94xx(F) aacraid(F) sx8(F) mpt3sas(OF) mpt2sas(OF) megaraid_sas(F) megaraid(F) megaraid_mbox(F) megaraid_mm(F) BusLogic(F) usb_storage xhci_hcd uhci_hcd ohci_hcd(F) ehci_pci(F) ehci_hcd(F) usbcore usb_common mv14xx(O) vgl(OF)
[    2.270007] CPU: 1 PID: 3972 Comm: insmod Tainted: GF          O 3.10.105 #24922
[    2.270007] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS rel-1.12.1-0-ga5cab58e9a3f-prebuilt.qemu.org 04/01/2014
[    2.270007] task: ffff88013a379800 ti: ffff8801367f4000 task.ti: ffff8801367f4000
[    2.270007] RIP: 0010:[<ffffffffa02feb80>]  [<ffffffffa02feb80>] virtscsi_queuecommand_single+0x10/0x30 [virtio_scsi]
[    2.270007] RSP: 0018:ffff8801367f7868  EFLAGS: 00010293
[    2.270007] RAX: ffff880138391000 RBX: ffff8801382e5d80 RCX: 0000000000000000
[    2.270007] RDX: ffff8801382e5d80 RSI: ffff8801382e5d80 RDI: ffff880138391000
[    2.270007] RBP: 0000000000000000 R08: ffff8801367f77a8 R09: ffff8801367f7798
[    2.270007] R10: ffff880136293d80 R11: 0000000000000000 R12: ffff880138391000
[    2.270007] R13: ffff88013610f7d8 R14: ffff88013610e848 R15: ffff8801382e5d80
[    2.270007] FS:  00007f3cea108700(0000) GS:ffff88013fc80000(0000) knlGS:0000000000000000
[    2.270007] CS:  0010 DS: 0000 ES: 0000 CR0: 000000008005003b
[    2.270007] CR2: 00000000000001d8 CR3: 00000001367ee000 CR4: 00000000000006e0
[    2.270007] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[    2.270007] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[    2.270007] Stack:
[    2.270007]  ffffffff8131866b ffff880138391000 ffff88013610e800 ffff880138391000
[    2.270007]  ffff880136808aa0 ffffffff8131e230 ffff88013610e970 ffffffff810b8283
[    2.270007]  ffffffff810cd816 ffff88013ffdd140 ffff88013610f7d8 0000000000000001
[    2.270007] Call Trace:
[    2.270007]  [<ffffffff8131866b>] ? scsi_dispatch_cmd+0xdb/0x370
[    2.270007]  [<ffffffff8131e230>] ? scsi_request_fn+0x2c0/0x4e0
[    2.270007]  [<ffffffff810b8283>] ? mempool_alloc+0x43/0x130
[    2.270007]  [<ffffffff810cd816>] ? pcpu_next_pop+0x36/0x50
[    2.270007]  [<ffffffff81259fca>] ? __blk_run_queue+0x2a/0x40
[    2.270007]  [<ffffffff81260505>] ? blk_execute_rq_nowait+0x65/0xf0
[    2.270007]  [<ffffffff812605e7>] ? blk_execute_rq+0x57/0xd0
[    2.270007]  [<ffffffff81260000>] ? blk_rq_map_user_iov+0xc0/0x160
[    2.270007]  [<ffffffff81131eb5>] ? bio_phys_segments+0x15/0x20
[    2.270007]  [<ffffffff812603fd>] ? blk_rq_map_kern+0xad/0x150
[    2.270007]  [<ffffffff8125c0f0>] ? blk_get_request+0x50/0xc0
[    2.270007]  [<ffffffff8131cfaf>] ? scsi_execute+0x11f/0x1d0
[    2.270007]  [<ffffffff8131e963>] ? scsi_execute_req_flags+0x83/0xe0
[    2.270007]  [<ffffffff8132158e>] ? scsi_probe_and_add_lun+0x1ee/0xf70
[    2.270007]  [<ffffffffa02fe405>] ? virtscsi_target_alloc+0x15/0x40 [virtio_scsi]
[    2.270007]  [<ffffffff81321046>] ? scsi_alloc_target+0x276/0x310
[    2.270007]  [<ffffffff81322677>] ? __scsi_scan_target+0xc7/0x5d0
[    2.270007]  [<ffffffff812758ce>] ? kobject_uevent_env+0x11e/0x4a0
[    2.270007]  [<ffffffff81301e6e>] ? get_device+0xe/0x20
[    2.270007]  [<ffffffff814ad60b>] ? klist_add_tail+0x1b/0x40
[    2.270007]  [<ffffffff813098b0>] ? transport_setup_classdev+0x20/0x20
[    2.270007]  [<ffffffff81322cb4>] ? scsi_scan_channel.part.12+0x54/0x70
[    2.270007]  [<ffffffff81322e31>] ? scsi_scan_host_selected+0xc1/0x150
[    2.270007]  [<ffffffffa02f26f0>] ? vp_interrupt+0x60/0x60 [virtio_pci]
[    2.270007]  [<ffffffffa02e51f4>] ? virtio_dev_probe+0x114/0x140 [virtio]
[    2.270007]  [<ffffffff81305eda>] ? really_probe+0x5a/0x220
[    2.270007]  [<ffffffff81306161>] ? __driver_attach+0x81/0x90
[    2.270007]  [<ffffffff813060e0>] ? __device_attach+0x40/0x40
[    2.270007]  [<ffffffff813041d3>] ? bus_for_each_dev+0x53/0x90
[    2.270007]  [<ffffffff81305678>] ? bus_add_driver+0x158/0x250
[    2.270007]  [<ffffffffa0302000>] ? 0xffffffffa0301fff
[    2.270007]  [<ffffffff81306768>] ? driver_register+0x68/0x150
[    2.270007]  [<ffffffffa0302000>] ? 0xffffffffa0301fff
[    2.270007]  [<ffffffffa0302081>] ? init+0x81/0xc7 [virtio_scsi]
[    2.270007]  [<ffffffffa0302000>] ? 0xffffffffa0301fff
[    2.270007]  [<ffffffff810003aa>] ? do_one_initcall+0xea/0x140
[    2.270007]  [<ffffffff8108bda4>] ? load_module+0x1a04/0x2120
[    2.270007]  [<ffffffff81088f60>] ? store_uevent+0x40/0x40
[    2.270007]  [<ffffffff8108c551>] ? SYSC_init_module+0x91/0xc0
[    2.270007]  [<ffffffff814c2dc4>] ? system_call_fastpath+0x22/0x27
[    2.270007]  [<ffffffff814c2d11>] ? system_call_after_swapgs+0xae/0x13f
[    2.270007] Code: 4c 8d 6c 0d 08 4c 89 6b 08 eb b2 29 d1 39 ca 77 e5 eb f8 0f 1f 84 00 00 00 00 00 48 8b 0e 48 89 f8 48 89 f2 48 8b 89 68 01 00 00 <48> 8b 89 d8 01 00 00 f0 ff 41 04 48 8d bf 08 05 00 00 48 8d b0 
[    2.270007] RIP  [<ffffffffa02feb80>] virtscsi_queuecommand_single+0x10/0x30 [virtio_scsi]
[    2.270007]  RSP <ffff8801367f7868>
[    2.270007] CR2: 00000000000001d8
[    2.340326] ---[ end trace fcf2542d630b56f6 ]---
 ... [FAILED]
 

 

Share this post


Link to post
Share on other sites

I get the following error when i use extra.lzma v0.6 on DSM 6.2.2 U5  918+ Juns 1.04b loader.

 

MB: Gigabyte H97N WiFi

CPU: Intel Xeon E3 1245 v3, 16GB Ram

Onboard NICs are working fine, but when i load the modules for Mellanox Connectx-3, i have a segmentation fault and the system crashes.

Any insights will be really appreciated.


[  163.275915] Compat-mlnx-ofed backport release: f36c870
[  163.275918] Backport based on mlnx_ofed/mlnx-ofa_kernel-4.0.git f36c870
[  163.275919] compat.git: mlnx_ofed/mlnx-ofa_kernel-4.0.git
[  163.286603] mlx4_core: Mellanox ConnectX core driver v2.2-1 (Feb, 2014)
[  163.286608] mlx4_core: Initializing 0000:01:00.0
[  170.240802] mlx4_core 0000:01:00.0: PCIe link speed is 8.0GT/s, device supports 8.0GT/s
[  170.240804] mlx4_core 0000:01:00.0: PCIe link width is x8, device supports x8
[  170.303681] mlx4_core 0000:01:00.0: Unsupported cable detected
[  170.328859] mlx4_en: Mellanox ConnectX HCA Ethernet driver v2.2-1 (Feb 2014)
[  170.329080] mlx4_en 0000:01:00.0: Activating port:1
[  170.332130] mlx4_en: 0000:01:00.0: Port 1: Using 64 TX rings
[  170.332131] mlx4_en: 0000:01:00.0: Port 1: Using 8 RX rings
[  170.332132] mlx4_en: 0000:01:00.0: Port 1:   frag:0 - size:1522 prefix:0 stride:1536
[  170.340165] mlx4_en: 0000:01:00.0: Port 1: Initializing port
[  170.340553] mlx4_en 0000:01:00.0: registered PHC clock
[  170.341243] general protection fault: 0000 [#1] PREEMPT SMP
[  170.346974] Modules linked in: mlx4_en(E+) mlx4_core(E) mlx_compat(OE) pci_stub vhost_net(O) kvm_intel kvm irqbypass vhost_scsi(O) vhost(O) ipt_MASQUERADE xt_REDIRECT nf_nat_masquerade_ipv4 xt_nat iptable_nat nf_nat_ipv4 nf_nat_redirect nf_nat tun cifs udf isofs loop tcm_loop(O) iscsi_target_mod(O) target_core_ep(O) target_core_multi_file(O) target_core_file(O) target_core_iblock(O) target_core_mod(O) syno_extent_pool(PO) rodsp_ep(O) hid_generic usbhid usblp uhci_hcd ehci_pci ehci_hcd nf_conntrack_ipv6 ip6table_filter ip6_tables xt_recent xt_iprange xt_limit xt_state xt_tcpudp xt_multiport xt_LOG nf_conntrack_ipv4 nf_defrag_ipv4 iptable_filter ip_tables x_tables btusb btrtl btintel btbcm hidp rfcomm bluetooth hid openvswitch gre nf_defrag_ipv6 nf_conntrack apollolake_synobios(PO) ax88179_178a(E)
[  170.419105]  usbnet tg3(E) cnic(E) bnx2(E) vmxnet3(E) pcnet32(E) e1000(E) sfc(E) netxen_nic(E) qlge(E) qlcnic(E) qla3xxx(E) sky2(E) skge(E) jme(E) uio(E) alx(E) atl1c(E) atl1e(E) atl1(E) libphy(E) mii(E) exfat(O) btrfs i915(E) drm_kms_helper(E) syscopyarea(E) sysfillrect(E) sysimgblt(E) fb_sys_fops(E) cfbfillrect(E) cfbcopyarea(E) cfbimgblt(E) drm(E) fb(E) fbdev intel_agp(E) intel_gtt(E) agpgart(E) video backlight button synoacl_vfs(PO) hfsplus md4 hmac r8168(OE) ixgbe(E) igb(E) i2c_algo_bit e1000e(E) vxlan ip6_udp_tunnel udp_tunnel fuse vfat fat crc32c_intel aesni_intel glue_helper lrw gf128mul ablk_helper arc4 cryptd ecryptfs sha256_generic ecb aes_x86_64 authenc des_generic ansi_cprng cts md5 cbc cpufreq_powersave cpufreq_performance acpi_cpufreq processor cpufreq_stats dm_snapshot dm_bufio crc_itu_t
[  170.491176]  crc_ccitt quota_v2 quota_tree psnap p8022 llc sit tunnel4 ip_tunnel ipv6 zram sg etxhci_hcd rtc_cmos(E) mdio(E) mpt3sas(E) raid_class(E) scsi_transport_sas(E) megaraid_sas(E) megaraid(E) mptctl(E) mptspi(E) mptscsih(E) mptbase(E) scsi_transport_spi(E) megaraid_mbox(E) megaraid_mm(E) vmw_pvscsi(E) BusLogic(E) usb_storage xhci_pci

xhci_hcd usbcore usb_common loalddpomlnp(OE) [last unloaded: apollolake_synobios]
[  170.528075] CPU: 0 PID: 4994 Comm: insmod Tainted: P           OE   4.4.59+ #24922
[  170.535768] Hardware name: Gigabyte Technology Co., Ltd. H97N-WIFI/H97N-WIFI, BIOS F4 06/28/2014
[  170.544689] task: ffff8802825f8180 ti: ffff8801e8a1c000 task.ti: ffff8801e8a1c000
[  170.552498] RIP: 0010:[<ffffffff8149545c>]  [<ffffffff8149545c>] dev_get_iflink+0x1c/0x30
[  170.560976] RSP: 0018:ffff8801e8a1fb78  EFLAGS: 00010202
[  170.566527] RAX: 655f34786c6d3601 RBX: 000000000000000c RCX: ffffea00089d715f
[  170.573943] RDX: 0000000000000801 RSI: 0000030600114bb3 RDI: ffff8801e8a60000
[  170.581188] RBP: ffff8801e8a1fb78 R08: 0000000000000000 R09: 0000000000000001
[  170.588641] R10: 0000000000000000 R11: 0000000000000000 R12: ffff8801e8a60000
[  170.595930] R13: ffffffff818620c0 R14: 0000010600114993 R15: 000001fe00114993
[  170.603342] FS:  00007f4cad2ee700(0000) GS:ffff88041ea00000(0000) knlGS:0000000000000000
[  170.611524] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[  170.617375] CR2: 000000000234a308 CR3: 0000000225414000 CR4: 00000000001626f0
[  170.624612] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[  170.632152] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[  170.639346] Stack:
[  170.641364]  ffff8801e8a1fb98 ffffffff814b1b43 ffff8801e8a60000 0000000000000000
[  170.648889]  ffff8801e8a1fba8 ffffffff814b1f09 ffff8801e8a1fbe8 ffffffff8149ff9c
[  170.656430]  ffff8801e8a60000 ffff8801e8a60000 ffff8801e8a60800 0000000000000001
[  170.664117] Call Trace:
[  170.666599]  [<ffffffff814b1b43>] rfc2863_policy+0x73/0x80
[  170.672282]  [<ffffffff814b1f09>] linkwatch_init_dev+0x19/0x20
[  170.678177]  [<ffffffff8149ff9c>] register_netdevice+0x23c/0x460
[  170.684246]  [<ffffffff814a01d5>] register_netdev+0x15/0x30
[  170.690159]  [<ffffffffa115c917>] mlx4_en_init_netdev+0x9b7/0xb80 [mlx4_en]
[  170.697368]  [<ffffffffa114d1af>] mlx4_en_activate+0xaf/0x100 [mlx4_en]
[  170.704166]  [<ffffffffa1112fec>] mlx4_add_device+0x8c/0xb0 [mlx4_core]
[  170.710828]  [<ffffffffa1165000>] ? 0xffffffffa1165000
[  170.716022]  [<ffffffffa11130a4>] mlx4_register_interface+0x94/0x100 [mlx4_core]
[  170.723753]  [<ffffffffa116509c>] mlx4_en_init+0x9c/0x9e [mlx4_en]
[  170.730407]  [<ffffffff810003b6>] do_one_initcall+0x86/0x1b0
[  170.736226]  [<ffffffff810e1b48>] do_init_module+0x56/0x1be
[  170.741987]  [<ffffffff810b7d8d>] load_module+0x1dfd/0x2080
[  170.747630]  [<ffffffff810b51f0>] ? __symbol_put+0x50/0x50
[  170.753184]  [<ffffffff810b8199>] SYSC_finit_module+0x79/0x80
[  170.759036]  [<ffffffff810b81b9>] SyS_finit_module+0x9/0x10
[  170.764689]  [<ffffffff8156a78a>] entry_SYSCALL_64_fastpath+0x1e/0x92
[  170.771186] Code: 45 31 ed 48 63 db 49 01 1c 24 e9 d6 fd ff ff 48 8b 87 00 02 00 00 48 85 c0 74 14 48 8b 80 d8 01 00 00 48 85 c0 74 08 55 48 89 e5 <ff> d0 5d c3 8b 87 20 01 00 00 c3 66 0f 1f 84 00 00 00 00 00 48
[  170.791322] RIP  [<ffffffff8149545c>] dev_get_iflink+0x1c/0x30
[  170.797380]  RSP <ffff8801e8a1fb78>
[  170.801060] ---[ end trace 85b6ec114d6b62f3 ]---

Share this post


Link to post
Share on other sites
On 1/22/2020 at 12:51 PM, vovetsp said:

Hello! Please add add realtek RTL8152/RTL8153/RTL8156 to driver extension for loader. https://github.com/bb-qq/r8152.

It works for on Asrock J4105b-itx.

 

thats driver version 2.12.0 from 4/2019 and already part of my extra.lzma, 3615/17 user can have it when using 0.6, only new 918+ is open as its still unreleased because of the problems with the i915 driver not working with all GPU gen's

 

Edited by IG-88

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.