IG-88

ich denke den kannst du in die tonne werfen, die alten 1068/1078 chips haben ein hardware limit von 2.3TB per platte pci karte gigabit mit realtek 8169, da gibt es zumindest einen treiber Intel Xeon E3-1225 v3, haswell 4th gen cpu mit gpu, die würde auch mit loader 1.04b 918+ laufen und würde hardware unterstütztes transcoding unterstützen dann lies mal das und lösche das betreffende verzeichnis, falls du doch mal auf 6.2 upgraden willst kann die das sonst auf die füße fallen https://xpenology.com/forum/topic/26723-network-down-after-dsm-login/?do=findComment&comment=137953 dann bleibt in der 6.1 linie mit dem loader den du hast aber ich würde auf die letzte 6.1 updaten um zumindest so viele sicherheitsprobleme wie möglich zu schließen, das wäre 6.1.7, das sollte keine extra probleme mit den installierten paketen machen nur wenn die cpu mit plex zu sehr ins schwitzen kommt und die die hardware der gpu nutzen willst (intel quick sync video), dann wäre aber die reinstalltion einiger pakete notwendig da du neben neuem loader auch eine migration auch einen anderen typ (3615 -> 918+) machen müsstest, ansonsten gibt bei 6.2.3 noch aktuelle sicherheitsupdates rein fon der funktion würdest du cloud anbindungen sicher zum arbeiten bekommen aber ich glaube nicht das sich das deswegen lohnt, im worst case musst du einige packete neu eirichten und das kann je nach art und dokumentation ein wenig arbeit sein nicht wirklich und nicht in deutsch, am nächten käme dem noch das hier, aber das stimmt nicht mehr ganz da es mitlerweile 6.2.3 gibt und man dafür dann das treiber pakete von 6.2.3 benutzen würde, außerdem setzt es das wissen zum normalen installieren aus dem tutorial bereich voraus (wie man den usb stick vorbereitet) https://xpenology.com/forum/topic/21663-driver-extension-jun-103b104b-for-dsm622-for-3615xs-3617xs-918/ so lange du das nicht wirklich brauchst lass es lieber, es kann dich einige stunden bis tage kosten,, da du sicer wieder einiges von der alten installation wieder auffrischen musst )die schritte bei 6.2 sind identisch nur nimmt man neuere loader und andere treiber pakete)

re check and also make sure there are no partitions on the disk you use to install

you should check the manual of your systemboards before buying m.2 sata ssd's, in most cases the sata part of the m.2 shares a sata connector of the board and when using the m.2 sata one of the sata ports is dead (when there is often a table that shows under what conditions of the m.2 slot what other port is dead and if there is a (good) logical picture of the board with chipset, pcie lanes, ..., then you might see a switch realted to the m.2 slot and the sata ports of the chipset

forget about SN and mac, for installing the usb vid/pid needs to match to the values of the usb flash drive you are using sn is only used for some packages and mac is only mandatory when using WOL

then yes, the new kernel of 6.2.3 uses a different method to backport the intel i915 driver and that makes loading of jun's driver impossible and because jun's driver is located in "update" it will be used instead of synologys driver (driver in "update" are supposed to be newer and overrule the drivers in the normal directory)

no the usb nic should not have to do with that [Wed Sep 9 01:14:14 2020] usbcore: registered new interface driver r8152 ... [Wed Sep 9 01:14:14 2020] r8152 2-1:1.0 eth1: v2.13.0 (2020/04/20) thats mainly the part that shows the driver is loaded and what driver it was the driver in the extra for 6.2.2 was 2.12 and in the newer for 6.2.3 it was 2.13, either 2.14 works or i will go back to 2.12 for 6.2.3

these are different drivers do you talk about a pcie card or a usb device from your (incomplete) dmesg file its r8152, so it should be a usb device if the r8152 driver in my extra/extra2 is not working i would prefer to give you a new test extra/extra2 wit a recent version (2.14.0) of that driver and you test it

if you cant use linux like tools an the bsd based macos you could use a virtual linux (https://download.virtualbox.org/virtualbox/6.1.16/VirtualBox-6.1.16-140961-OSX.dmg) https://xpenology.com/forum/topic/25833-tutorial-use-linux-to-create-bootable-xpenology-usb/ but maybe you dont need that at all, if you write the img file directly to usb and edit the files there? osx should be able to use usb flsh drive in FAT format (i guess)

yes but without amd support in kernel of the three types we have loaders for the amd support in vmm might not be usable

what loader and dsm type did you use? 1.04b 918+ will not work, i has to be 1.03b, you might try 3615 you dont need there two for just booting and finding it in network,, vid/pid is needed when installing dsm to disk and mac only when using WOL

its possible to make that, it does not exist yet jun's extra.lzma does not contain a driver r8125.ko, so if you are looking for a better driver you won't find it that way does the system boot when you remove the 8125 card?

no intel gpu support, 3615/17 dont have i915 drivers and 918+ needs min. haswell (4th gen) cpu https://xpenology.com/forum/topic/13333-tutorialreference-6x-loaders-and-platforms/

may be you should read what i wrote in the 6.2.3 post about that matter, it exists since 6.2.0 and happens with jun's original drivers, you can/will have the same problems with 6.2.2 jun did not document how he made the drivers and from what kernel source it was discovered pretty late and seems not to effect a lot of people or we would have seen complains earlier its not so difficult now to get around that by using ahci controllers (i do it that way, also the control over the write cache of the disks does not work with lsi sas controlelrs but does with ahci)

good to know that it's solved and there is no unknown behavior, i guess the two usb ports where just virtual ports and for that reason not interacting with your usb drives

the messages you see with xpenology are from juns loader, you might see nothing with a normal original loader, you might see something with a serial console and even if it boots it might be a short process and might stop shortly after starting the kernel and there might never anything to find in network as the nic driver is missing or the protection kicks in because some pci id's are missing (that would be present on the original hardware) imho you cant say that for sure with just that little test i don't know how deep you looked into the atom cpu (intel) desaster, it was a degrading process and even the resistor was only slowing it down, all damage already done could not be reversed and if a cpu started to have problems there was slight chance it might come up again with the fix but in that degraded state its likely to fail later mine never woke up again and the whole system board was useless as the cpu could not be changed (new system has a socket again)

you were not specific about the card but the driver i'm talking about are already part of the loader and will be started then booting i tried to explain above, maybe just for fun go into the controller bios create a raid set of one or more drives and then see if something new turns up in dsm, the way you want/expect is is different and thats called IT mode with the lsi sas controllers that's because you don't need to care about that, maybe if your driver is not part of the default set, but there is a extended set already here https://xpenology.com/forum/forum/91-additional-compiled-modules/ but maybe try the default set for now before making it more complicated then it needs to be maybe read this about IR and IT mode of lsi sas controllers ans start, gives you a basic view what it is about IT mode and how to do it (you would need to check if there is a way to check your specific card as it needs the right firmware) https://flemmingss.com/how-to-flash-it-mode-firmware-to-hp-lsi-sas-9212-4i-controller-card/ https://www.reddit.com/r/DataHoarder/comments/6t1vz5/do_i_need_to_flash_my_lsi_sas_card_to_it_mode/

nothing unusual for usb (2 usb ports detected) there is still another option if usb works with the default config the sata controller has 6 ports but only uses 2 ports dir disks, if you can change this in the vm config to two ports as needed then your sas controller will get lower and you will be within the 12 port limit of the default config it can also be tried to add a different (usb3) controller in the vm config or trying with more usb ports and assign the hardware device (usb disk) to a higher usb port

bigger udpates/upgrades reset the synoinfo.conf and its way beyond what the system is made for to use usb as internal drives, can break with every update, synology does not test or care about this kind of scenario, risky thing to do as you found out and i'm sure you have been warned here to do this i guess you need to check your synoinfo.conf and redo your configuration change the its not working in the way i expected it, maybe send a dmesg log you could try to define less internal disks and have usb start lower maxdisks="16" esataportcfg="0x0" usbportcfg="0xffff0000" internalportcfg="0xffff"

maybe start reading this and adjust the plan? https://xpenology.com/forum/topic/13333-tutorialreference-6x-loaders-and-platforms/ even 3617 only supports max 16 cores (ht included) so with 2 x 8core+HT it would be 32, i'd suggest diabling HT and the its 16 cores (when counting 1 x HT as 25% of a real core you loose the performance of 4 cores) remember the faq? https://xpenology.com/forum/topic/9393-hardware-faq/?tab=comments#comment-82386 the concept of dsm is pure software mdadm raid, there are added drivers for raid controllers but you would need to define a raid in the controller to see the volume as disk in dsm it can be one raid0 for every single disk or a bigger raid5/6 set, in both cases you have not much control as you dont have the controllers tools to manage the raids (dsm is a closed source appliance) with a lsi megaraid you should check if there is a IT firmware for it and use this one, called HBA mode with other controllers, every disk will be seen as single drive on a original unit? xpenology can do more but also sticks to some limits as its still using the same base kernel and dsm base system, also its bound to some models that are usable with the loaders

as you only added 2 ports (and usual system have much more usb ports) they might simple be beyond the default value of 2 ports, kind of same thing as with the sata ports, if its only 12 and you have ports beyond that you will not see them just make it usbportcfg="0xffff00000" and see what happens

did you load all in t the right order? when loading the modules in the right order (dependencies) then this should be no problem, my error was a different one i guess it should be like this? insmod /usr/lib/modules/cfg80211.ko insmod /usr/lib/modules/mac80211.ko insmod /usr/lib/modules/update/iwlwifi.ko insmod /usr/lib/modules/update/iwlmvm.ko insmod /usr/lib/modules/update/iwldvm.ko but won't we need the driver 1st in dsm to be working? these are VAAPI devices, i have not seen them when loading the nvidia drivers after insmod /urs/lib/modules/nvidia.ko insmod /urs/lib/modules/nvidia-uvm insmod /urs/lib/modules/nvidia-drm /usr/bin/nvidia-smi -pm 1 i only got /dev/nvidia0 /dev/nvidiactl i compiled the nvidia kernel drivers by unpacking the file from nvidia got int /NVIDIA-Linux-x86_64-440.44/kernel (i did not put anything into the kernel source) and did the normal cross compile in my chmod environment inside a dsm vm make ARCH=x86_64 CROSS_COMPILE=/test4/x86_64-pc-linux-gnu/bin/x86_64-pc-linux-gnu-

i had a opportunity to test these card/chip and it does not work as serial console with both kernels

the other two drives are useless for the raid1 of the system if you already have /dev/md[x] you can mount the /dev/md... of these three sd[x]1 partitions and make your change in the synoinfo.conf maybe 1st restore the default you had before, test it by booting the vm, look in the gui if there are any things to fix with the raid (if 2 disks having a re volume are not part of the system raid1 there should be something to repair if everythin is back to normal you can change to what i suggested above

maybe the just copy the file to a share the nas provides, login as admin by usins ssh with puty type "sudo su" and hit return, that will result to work the same as root, now you can change permissions of a file and copy it where you need it and mind the words you see when you do the sudo su the first time

the mac of the system the nas is running on, its important if you want to use WOL yes and no, you will need it every time the system boots so it needs to be where it is you can remove it while its running for backup or updating it (rarely needed), its not mounted when the nas is running looks like i forgot to mention the faq https://xpenology.com/forum/forum/83-faq-start-here/