IG-88

if you have boot problems then 3615 and 3617 are the same, you can stick to your former 3617 and test with this loader and extra.lzma start with just the plain loader, no changes to any files, the vid/pid of the usb will get important when starting to install DSM, just for finding it in network vid/pid is of no importance if you solved this problem you can start modding the grub.cfg with vid/pid and mac (mandatory when using WOL) as documented here https://xpenology.com/forum/topic/13333-tutorialreference-6x-loaders-and-platforms/ 1.03b/DSM 6.2 does not support uefi anymore, only CSM/legacy mode so look in bios for that option and its also important have a look at the boot selection, often there are uefi usb devices and the same usb turns also up as legacie device (maybe without uefi in its display name) after you enabled CSM/legacy mode you need to save and reboot, then connect your xpenology usb before going into bios!!! or you might not see the legacy usb device in some cases legacy mode can be exclusive and you dent see uefi recources but in most cases its more like side by side and you see uefi boot devices and legacy devices in boot selection with your hardware you should not need a extra.lzma when using dsm 6.2.0 (23824) or DSM 6.2.3 (25423), the driver jun provides in his default cover the realtek nic when using 6.2.2 (24992) you will nee a different extra.lzma from here https://xpenology.com/forum/topic/21663-driver-extension-jun-103b104b-for-dsm622-for-3615xs-3617xs-918/

thats not going to happen, your raid of the data volume is not touched when doing system updates in worst case you just loose the ability to boot DSM, you can always another OS to access your data on the raid volume my plan B is still to start open media vault and let it discover my DSM volume (dsm uses just the normal linux features like mdadm and lvm2 any linux system can use) you can prepare a usb drive or disk with OMV (disconnect any dsm drive/usb prepare omv on your hardware) and boot from it instead of booting from the xpenology loader it should see your raid volume and you cam make is accessible through network for windows or mac, omv has i nice and easy web gui too as long as you dont need dsm specific features you can use OMV until you can make DSM boot again there are other option too (see downgrade tutorial with a additional disk) and even deleting the content of the 1st(!!!) partitions of your DSM drives and reinstalling DSM from scratch (even with a older dsm version), dsm will not touch your raid data partitions, it would only reclaim the 2.4GB system partitions, there is a mode to overwrite the all data including your data partitions but you explicitly need to choose that and acknowledge that you loose all data (taking the time to read a warning is not vasted time) a semi bricked dsm is no big deal if you are prepared and understand your system (as with any self installed server system) and having a backup is also a good idea because even raid6 can break and file system can have problems too (even if the raid works fine) people not wanting to deal with this stuff are better of buying a real synology box or a qnap box (and still should have a backup at least from the most important data like family photos or important documents - separating data under that point of view will make things easier to backup, in most cases a big usb disk is enough for having a backup of the most important stuff)

als mit dem hast du ganz sicher kein hardware transcoding, das "F" steht für "ohne GPU" und intel quick sync video ist teil der iGPU https://ark.intel.com/content/www/us/en/ark/products/190886/intel-core-i3-9100f-processor-6m-cache-up-to-4-20-ghz.html ich habe mir vor kurzem eine i3 9100 gekauft und wollte die jetzt mal mit 6.2.3 testen 6.2.3 hat neue i915 treiber direkt von synology, jun's treiber die 1/2 jahr neueren sourcecode verwenden und für 6.2.0-6.2.2 gehen laufen mit 6.2.3 nicht mehr, wenn man also aktuelles DSM nutzen will dann ist kompatibel mit 6.2.3 angesagt i3 8100 wäre safe bet, die geht auf jeden fall hängt sehr davon ab was man für ein gehäuse hat und ob/wie man später skallieren will ich habe microATX mit geringerer tiefe (fast so gering wie miniITX), hat mehr pcie slots zum erweitern (zusätzliche ahci controller, 10Gbit netzwerk) optimal für mich war gigabyte 365M HD3, geringe tiefe und 4 pcie slots (je einmal 8x und 4x war muss, 1x eher wenig nützliche zugabe) wer unbedingt nvme lese/schreib cache haben muss der braucht 2 x M.2 nvme von dem was nachher an performance als nas so rüberkomt dürfte es egal sein ob du B360 nimmst oder Z390, wichtiger kann sein ob man ECC ram will oder nicht, das wirkt sich dann auch cpu/chipsatz aus ich würde aber mal von meinem favoriten abgesehen eher zum GIGABYTE Z390 M raten, lässt mehr optionen offen ist zwar viel benutzt aber nur eine pcie lane und PCIe 2.0 -> max. 500MByte/s das ist für 4 disks ein wenig magerm besser wäre 9235 basierend das sind wenigstens 2 lanes (braucht dann aber auch einen "größeren" pcie slot als 1x) oder mein neuer favorit JMB585 (habe ich heute geliefert bekommen), 5ports, 2 lanes aber PCIe 3.0, hätte bei 5 statt 4 ports einen 4fachen durchsatz am pcie bus außerdem kann man einen M.2 nvme slot mit adpter auch für eine pcie karte entfremden, z.b. 10Gbit nic und es gibt auch M.2 sata controller (z.b. besagten JMB585 als M.2 mit 5 Ports), aber auch ein 8port ahci controller mit 2 pcie lanes ist machbar - habe zwei M.2 -> PCIe 4x Slot Adapter hier liegen aber noch nicht getestet (war zusammen mit dem i3 9100 neubau geplant) ok ich bin aus der silent fraktion, mein nas steht im wohnzimmer, da kommt so ein 19" nicht in frage weil zu tief und zu laut 4 platten wären mir auch zu mager, ich habe jetzt 12 und beim nächten gehäuse mit größeren platten wären es vermultich 6-8 slots (aber ich habe mich wegen der 12 von hotswap verabschiedet)

sollte er eigentlich nicht, aber stört erst mal nicht, wenn man sich die grub.conf kopiert und weiß welche dsm version man hatte kann an den usb jederzeit neu erzeugen (falls man ihn versehentlich löscht), oder man macht sich einfach eine kopie mit "Win32DiskImager 1.0"

if you read this https://xpenology.com/forum/topic/13333-tutorialreference-6x-loaders-and-platforms/ you might recognize "... AMD Piledriver is suspected to be the minimum chip architecture to support DS916 or DS918 DSM platforms, but this is unverified. ..." so even "older" amd cpu's are capable of running 918+ and its also interesting to see what hardware people use and make assumptions (like using 918+ with older xeon cpu's that come without a iGPU) https://xpenology.com/forum/forum/78-dsm-updates-reporting/ if anyone is interested about the cause, synology also published the kernel config used i don't know which of these parameters make a 4th gen intel cpu mandatory (its alos clear to see that 8 cpu cores are a hardcoded limit within the 918+ kernel, CONFIG_NR_CPUS=8, so 4 core with HT or 8 without HT, 3617 has 16 in that place)

DSM writes the kernel files (zImage and rd.gz) to the usb - if there is a new one with the update - to "revert" you just replace these files with your old you had before the update you can also use Win32DiskImager 1.0 to make a copy of your usb before doing the update (activate "read only allocated partitions" to keep it fast an small), if Win10deas not get you a drive letter anymore (1909, latest udpates) to use the tool you can make one with "Ext2 Volume Manager" (https://www.ext2fsd.com) use the 2nd partition that, same tool/way can used to copy end revert the two kernel files btw information's about the installed version are stored in the file /etc/VERSION /etc.default/VERSION same file is also in the rd.gz so on boot DSM will compare if there is a mismatch also this file is in the *.pat file used for updating, at least in not "updateX" updates (means biger updates that change at least minor version numbers) in samller "updateX" updates there can be a new kernel too, in this case its inside "flashupdate_6.2-24922-sX_all.deb", these are the files that are written to your usb (zImage and rd.gz. will be there too) example: majorversion="6" minorversion="2" productversion="6.2.2" buildphase="GM" buildnumber="24922" smallfixnumber="5" "smallfixnumber" would be for a "updateX" version so the 5 in the example is the red X from above updateX version always contain the content of former updateX versions so update6 will have all files from update1-5 too (can be checkr be opening the *.pat like a zip file)

looks like i will do it this weekend, today i got my two JMB585 based, pcie 3.0, 5port ahci controllers (one pcie 4x, one M.2)

in most cases its about csm/legacy, it needs csm mode in bios and to use the legacy usb boot device 1.03b/dsm6.2 does not support uefi https://xpenology.com/forum/topic/13333-tutorialreference-6x-loaders-and-platforms/ but 918+ might be a better choice anyway, should work and even the gpu should be usable for hardware transcoding (when its the same cpu as in the 1st post) for transcoding (beside having /dev/dri after booting up) https://xpenology.com/forum/topic/24864-transcoding-without-a-valid-serial-number/

also wenn der 1.02b ohne mbr geht dann brauchts das imho auch nicht für 1.03b bleibt also als unterschied das bei 1.03b kein uefi unterstützt wird, als noch mal biso wegen csm checken und sicher stellen das auch vom csm/legacy usb boot device gestartet wird, man kann bei vilen systemen csm aktiv haben und dann doch mit dem uefi boot device im uefi modus starten, dsm ist da nur eine option die nur greift wenn man auch vom csm/legacy device bootet

no, not anymore https://xpenology.com/forum/topic/24864-transcoding-without-a-valid-serial-number/ can you PM me a dmesg log? (/var/log/dmesg)

in der sektion "Member Tweaked Loaders" gibts einen thread https://xpenology.com/forum/topic/20270-dsm-62-loader-with-mbr-partition-table/ und da drin ist im ersten post ein link zu dem loader außerdem meine ich mich zu erinnern das mal jemand eine anleitung gepostet hat wie man den loader zu mbr wandelt

any source code for a driver you can provide as a link? do you have the pci vendor/device id of that card like [1924:0813] you can see with lspci or in dmesg log? shortly had someone asking about a solarflare nic https://xpenology.com/forum/topic/21663-driver-extension-jun-103b104b-for-dsm622-for-3615xs-3617xs-918/?do=findComment&comment=142932 it was for a solarflare sfn5162f and it looked like it might be supported already, yours might be older so it might work too from reading a few specs is seems to use the SFC9020 controller and that would be pci id [1924:0803] and this id is in the driver sfc.ko of the 0.5 extra.lzma for 3615 and 3617

yes i know, i referenced to the x20+ units yesterday in my 1st post, that's the units where synology did not retract the firmware as they need it to work (gemini lake support starting with dsm 6.2.3 aka new i915 driver), there where already signs for coffee lake and gemini lake support since last year in the source code from synology where does it help with the problem of your i5 8500t, asus h370i gaming and 6.2.3? having a log or error message might help to understand why video station does not work on your unit (btw. does it crash or simply does not offer hw transcoding, or ...) not really, it was checking cpu and board supporting iGPU and seeing that the driver is not loading because of the fact that there is the device missing just reading the manual from1st to last page would not have solved it, you would need to know that the internal gpu can be disabled (usually the case when a normal more powerful gpu is present)

thanks, at least we have information's any logs from 6.2.3 aka synology driver for ds video hardware transcoding attempt? beside the driver code there is only one firmware file different, "cnl_dmc_ver1_06.bin" vs "cnl_dmc_ver1_07.bin" to juns driver and cnl is for cannon lake, so i guess it does not make a difference for a coffee lake gpu, i guess as coffee lake is gpu gen 9.5 its still kbl firmware for coffee lake at least there is a solution, stay with 6.2.2 (jun's i915 driver)

i have a i3 9100 cpu for my new build, i guess if 630 is to blame i will now in the next xx day's :-) afair i3 8100 (also 630) was reported to work

https://www.backblaze.com/b2/storage-pod.html and if you use sas you can have so much more drives per controller its more a question of power consumption, noise and money

any logs? does not hurt to cross check with the patching tool, it creates backup copy's and it can be reverted did it work with jun#s i915 drivers on 6.2.0 / 6.2.1 / 6.2.2 ?

dont be shy :-) from the change log its just about universal search (you would need to update that plugin to 1.5 before updating to 6.2.3) i did not have any problems when testing with 6.2.3/918+ tbh my "productive" system still runs on 6.1 and is waiting for it's new hardware to be assembled - but i would use 918+ 6.2.3 if i update - still plan to do this an a daily base ;-)

changed that lately and dont have any problem with osfmount maybe it makes a difference when the usb drive was used on a mac or liunux system?

any logs or errors? what did you use to make sure you dont have a license problem? this is working for sure https://xpenology.com/forum/topic/24864-transcoding-without-a-valid-serial-number/

thats definitely not going to work as its "only" a 3rd gen processor (i7-3770) 3615/17 1.03b need CSM mode with UEFI bios and the legacy usb boot device, using loader 1.02b (3615/17) is not that picky and will work with both so its a good choice for at least testing and as there is also a mbr version (i used it for some hp desktops) its not to complex to test, just two versions 1.02b normal and 1.02b mbr to make sure the intel nic is supported i'd suggest using my extra.lzma v4.8 https://xpenology.com/forum/topic/9508-driver-extension-jun-102bdsm61x-for-3615xs-3617xs-916/

??? der default im 1.04b loader ist 16 da muss man nichts erweitern wenn 6 x sata gefunden wird (obwohl man nur 4 auf dem board hat - habe ich hier bei einem board auch) und man einen 8 port sas controller dazusteckt (bei 3615/17 schon da ist 12 der default) hier mal zwei schnipsel aus dem patch von des 1.04b loaders diff --git a/etc/synoinfo.conf b/etc/synoinfo.conf index 91cc8dd..2a4a12e 100755 --- a/etc/synoinfo.conf +++ b/etc/synoinfo.conf @@ -136,7 +136,7 @@ enableRCPower="no" maxaccounts="2048" maxgroups="256" maxshares="512" -maxdisks="4" +maxdisks="16" maxprinters="2" maxlogsize="64" @@ -263,8 +263,7 @@ upnpmodelurl="" upnpmodelname="DS918+" sdcopy="no" upnpmanufacturerurl="http://www.synology.com/" -internalportcfg="0xf" -HddHotplug="yes" +internalportcfg="0xffff" support_ssd_trim="yes" support_fan_adjust_dual_mode="yes" max_volumes="512" ich hätte wohl bis zum schluss lesen sollen was auch immer das für probleme sind mit dem mapping habe die ziemlich sicher nichts zu tun das dmesg log hilft hier nicht, das normale log sollte die fehler/probleme zeigen und evtl. auch mal bei samrt/health in der gui reinschauen ich hatte schon probleme mit seagate platten am lsi (bisher nur hgst) und habe die dann auf die internen ports geklemmt ansonsten sehen das man netzeil probleme ausschließt (weniger platten im probebetrieb, anderes netzteil) und sehen ob man es an einem der 4er kabel des sas controllers festmachen kann (nur 4 platten an je einem kabel) oder ob man es auch onboard vs lsi controller eingrenzen kann jeweils 4 platen klingt gut, also 4 am internen sata, danach 4 am 1. sas port und dann 4 am 2. sas port auf jeden fall das log checken !!!

from the dmesg log's pci device list there is no graphics device from the cpu if there is not hardware device visible to the operating system it cant use a driver, you might study the manual (any gpu disable functionalitys in bios?) or consult the supermicro support why there is no such device here what i mapped as 8086 vendor id [8086:a12f] type 00 class 0x0c0330 100 Series/C230 Series Chipset Family USB 3.0 xHCI Controller [8086:a131] type 00 class 0x118000 100 Series/C230 Series Chipset Family Thermal Subsystem [8086:a13a] type 00 class 0x078000 100 Series/C230 Series Chipset Family MEI Controller #1 [8086:a13d] type 00 class 0x070002 100 Series/C230 Series Chipset Family KT Redirection [8086:a102] type 00 class 0x010601 Q170/Q150/B150/H170/H110/Z170/CM236 Chipset SATA Controller [AHCI Mode] [8086:a110] type 01 class 0x060400 100 Series/C230 Series Chipset Family PCI Express Root Port #1 [8086:a114] type 01 class 0x060400 100 Series/C230 Series Chipset Family PCI Express Root Port #5 [8086:a118] type 01 class 0x060400 100 Series/C230 Series Chipset Family PCI Express Root Port #9 [8086:a119] type 01 class 0x060400 100 Series/C230 Series Chipset Family PCI Express Root Port #10 [8086:a11a] type 01 class 0x060400 100 Series/C230 Series Chipset Family PCI Express Root Port #11 [8086:a149] type 00 class 0x060100 C236 Chipset LPC/eSPI Controller [8086:a121] type 00 class 0x058000 100 Series/C230 Series Chipset Family Power Management Controller [8086:a123] type 00 class 0x0c0500 100 Series/C230 Series Chipset Family SMBus [8086:1533] type 00 class 0x020000 I210 Gigabit Network Connection [8086:1563] type 00 class 0x020000 Ethernet Controller 10G X550T [8086:1563] type 00 class 0x020000 Ethernet Controller 10G X550T [8086:15b7] type 00 class 0x020000 Ethernet Connection (2) I219-LM [8086:1918] type 00 class 0x060000 Xeon E3-1200 v5/E3-1500 v5/6th Gen Core Processor Host Bridge/DRAM Registers the device we would expect [8086:191d] HD Graphics P530 source: https://ark.intel.com/content/www/us/en/ark/products/88173/intel-xeon-processor-e3-1245-v5-8m-cache-3-50-ghz.html https://pci-ids.ucw.cz/read/PC/8086/191d edit: according to the manual you could check in bios for Graphics Configuration Primary Display -> IGFX Internal Graphics -> Enable (i'd expect it to work ootb when no extra graphics card installed but it cant hurt to try the settings above) is a monitor connected to a video out of the board and is working?

yes 88SE9215 chip is a 4port controller (but has one one pcie lane) yes, from my link above about asm chips you can see its a 2port controller so you should be able to use two ports

mine looks like this and make sure to unmount before installing mkdir -p /mnt/synoboot1 mkdir -p /mnt/synoboot2 echo 1 > /proc/sys/kernel/syno_install_flag mount /dev/synoboot1 /mnt/synoboot1 mount /dev/synoboot2 /mnt/synoboot2 yes when using 6.2.2 it should be 0.8 standard for 918+ the cpu should support it