IG-88

3615 image might boot up on any 64bit intel cpu (amd can be a problem), if its driver problem you can try the driver extension (extra.lzma) if you are more specific we could tell you more, so its just crystal ball gazing and maybe some occlumency https://xpenology.com/forum/topic/9508-driver-extension-jun-102bdsm61x-for-3615xs-3617xs-916/

sounds like this hardware problem (i have a asrock C2550D4I) https://www.anandtech.com/show/11110/semi-critical-intel-atom-c2000-flaw-discovered afair asrock at least extended the warranty to 3 years, so if you bought it 2015 you might give asrock support a call and get a free replacement https://www.auvik.com/media/blog/vendors-clock-signal-flaw-intel-atom-c2000/ in December 2016, user ScHAmPi posted this response from ASRock support in a forum thread: “Thank you for reaching out to us with your concern! Your Server board is using Intel Avoton CPU which is also listed in Intel’s errata announcement. … We provide 3-year manufacture warranty; if you ever experienced a sudden No POST with your board, please contact us. We can run a simple test to find the telltale sign. If the issue is confirmed that it may be related to the CPU, we will help you replace the board under our warranty policy.” https://www.reddit.com/r/freenas/comments/4x1kh1/asrock_c2550d4i_sudden_death/

just disconnect the keyboard and try the shutdown and start again in addition you can enable the bios option that the system always starts when power is connected (usualy its set to "last state")

das mit der 10s reparatur ist beim bootloader, wenn da etwas angepasst werden muss (zu alter loader auf usb gegenüber der platte mit neuerem system, dann wird der loader von derplatte auf den stick geschrieben und neu gebootet) das eine platte aus dem raid "fällt" kann schon passieren wenn man einmal ohne diese platte startet dann ist das raid nicht mehr synchron und die platte muss neu intergiert werden, das dauer dann etwas, das raid ist ja auch ohne die 3. platte funktionsfähig und wird beschrieben, imho kennt das raid kein transaktionslog (das man nutzen könnte um alles nach transaktion x nachzuholen) so das alle sektoren der platte überprüft/synchronisiert werden müssen (bzw. eher die clustergröße im raid die mit prüfsummen geschützt ist)

hast du den 3615 usb stick auch neu geschrieben (mit der org. "alten"version von jun)? wenn du nach dem wechsel auf 3617 wieder den alten stick nimmst (ohne ihn neu zu bespielen) ist da immer noch der neuen kernel von dsm 6.2 drauf, bei den updates wird auch immer der kernel auf dem usb stick mit aktualisiert und beim versehentlichen update der 3615 auf dsm 6.2 ist eben auch der neue kernel auf dem usb stick gleandet und wen du dann anfängst zu booten wird der (zu) neue kernel vom usb gestartet und man sieht im syno assistant wieder die neue kernel version von dsm 6.2

i've re-copied the original apollolake config and recompliled mvsas and libsas with CONFIG_SCSI_SAS_ATA=Y try this 0.5_test version http://s000.tinyupload.com/?file_id=01077994252925895830

usually not, never had problems with that as a solution you should check /var/log/dmesg for messages from the mvsas driver and disks that are found dsm counts the sata ports if the ports exceeds the defines max. 12 ports (even unused ports) disks can be present in the system (linux) but will not show up in the dsm gui, so if you check look for disks, count them and look for vendor id's of your missing disks (just ssh into your system with putty type dmesg and you can scroll up/down in the window, or use a editor to check the log)

they have a bios too, maybe you PM him to ask for any unusual he might have seen with his setup that 1.02 does work is only important for the fact that the hardware is working in general, as 918+ is using different kernel settings and newer kernel 4.4 there are a lot of differences (min cpu spec. as one, and i've seen the legacy mode option to make trouble on my test system) if you connect vga and see the screen go dark/blank after jun's message you can be sure it crashed the kernel on boot completely there is another hing you can try, install 918+ on a different hardware (single disk) and use this usb and hdd on your dell, after it booted and a few minutes without hdd activity switch off, boot usb live linux (or connect hdd somewhere else) and check if there is something new in /var/log/dmesg, if the system came up without a proper nic driver you should see that, if there is nothing new in dmesg then the switch from usb boot to hdd did not work, its likely that the system crashed before that (the hdd controller seemed to be ahci so storage driver should not be a problem in this test)

1st post (from @jun) in this thread, 3rd word "here", thats the link @CrazyCreator posted

sudo su and after that you have root privileges an can change what you want but be carefull its "you ask for it you got it" no safeties to prevent you from doing stuff to mess up the system this controller should use megaraid_sas.ko as driver, thats part of juns loader and it should work ootb (in theory) without additional driver package

as its NVMe and thats not supported on that systems (3615/3617) there is not much, i've compiled a nvme driver but as synology had its own ways there where lots of things missing that where present on the newer 918+, so the one or two people thinking about using it had given up (with the nvme driver the disk device is at least created if you check dmesg) one alternative is to use esxi (virtual xpenology), format the nvme as vmfs, create a virtual ssd and hand that over to the xpenology vm not sure that is a good choice and its some way from baremetal to esxi, if you familiar with esxi then it might be ok but if not you are up to a lot of work to catch up on that (pass through or not, ...) - not woth it imho if its just for xpenology 2nd option available now is the new (alpha/testing) loader 1.03a2 with the 918+ image, that has native nvme support build in by synology and might be able to use the nvme ssd (you can test by disconnecting all disk and use a single empty disk for a test install) if you want to use it as cache then the performance gain by just one ssd (as read cache) is extremely small (or none) even with two and write enabled you will not see much, its a option that will pay off when having lots of users accessing the system over network

did you see my comments on the bottom of the 1st posting here mainly 3. about bios setting and if you can use a serial cable you might see if the system crashes on boot or just have driver problems the nic on your dell is a intel 82579, thats e1000e driver, my extra.lzma contains the driver from the april 2018 driver intel driver cd 23.2 and i tested/used that driver on my test system (baremetal) i'd say its unlikely a nic problem you can try with jun's loader without additional extra.lzma on my test system i could recognize a crash (legacy bios option) by seen the vga monitor going blank instead of showing jun's message, in that case it was about the i915 driver (graphics) but you cpu does not have graphics so also unlikely the entry in my cpu list if from that:

good to hear if you have some time, can you check the loader 1.03a2 and my extra.lzma 0.4 for 918+ image works with your asus xg-c100c nic? its the same source code version in there but there could be problems with kernel 4.4 that 918+ uses (could not get my tehuti 10G nic to work properly and some intel drivers) some feedback on that would be nice too https://mega.nz/#F!BtFQ2DgC!JgomNP3X8V9EuwxL4TXbng https://xpenology.com/forum/topic/12859-driver-extension-jun-103a2dsm62x-for-ds918/

the driver in the v4.8 extra.lzma is not recent anymore, maybe a newer driver helps i've compiled aquantia x2.0.10.0 driver for 3615 so you can test that v4.9_test version in the following link (only new compared to 4.8 ist the aquantia driver) http://s000.tinyupload.com/?file_id=05787267927444477854

oops, wrong name, thanks, its synoinfo.conf and there is one in /etc and one in /etc.defaults to still have the change after a reboot you will need the edit in /etc.defaults too changed it to "in synoinfo.conf (in /etc AND /etc.defaults)" change it in /etc.defaults/synoinfo.conf and reboot to get the rocketraid working you will need the driver mvsas.ko, that one is part of my driver extension (-> STORAGE DRIVER LIST, Marvell)

i have downgraded the ixgbe driver to the first version that would support your nic, you can try the new version 0.4 from the 1.03 extra.lzma thread

there will be no new extra.lzma after 0.4 for the next 5 weeks, please leave feedback about the driver but don't expect answers or new versions

i've redone the mpt2sas driver and got rid of the this message tested hot plugging, created a volume and unplugging a disk from the LSI 9211-8i, at least with the driver as its now it worked (did not tested it with the driver in the 0.2 extra.lzma) with the new driver that does not work the synology way, diskmanager does not show the correct values but "smartctl" does show correct values when called on the command line so the driver mpt2sas is working properly, only dsm fails to use the information, i guess synology has its ways to filter and the information gets lost on the way (with 918+ its not as with the business units 3615/3617 those come natively with lsi/sas support) for now its not a driver thing, smart does work as seen in the output, maybe some else might have the time to look into this to see if it can be fixed by addind some scripts or programs from the business units or adapt things from there i've test this too with the driver that will be in extra.lzma 0.3 (will be later this evening online), it did work without problems connected a disk already installed on a ahci port with dsm as sigle disk to the lsi, came up without problem emptied a disk and connected it to the lsi (2 port ahci + 8 port lsi) as disk 8 and booted up the the usb flash drive, installed to the disk without problems and booted after installed edit: looks like the reboot only happens when more then a system/swap partition is on the disk, reproduced the reboot with a SHR volume on the 2nd disk also when reconnecting the disk is not recognized as formerly used disk, only after reboot the system partition and SHR partiton was recognised so replugging is also not really working hotplugging should be avoided when using the mpt2sas driver so the driver is in basic working condition from my point of view (cant see how to improve it further) won't get better as i cant find out why the newer driver mpt3sas (that also support mpt2sas controllers) is not working - see my post above and the spoiler part with the log if someone has a idea whats wrong drop a comment here (i've tried mpt3sas from the kernel and v23 from external source, both have the same crash)

i'm pretty sure @Polanskiman will do that later, the 6.1 tutorial dates mid september 2017 so don't expect a 6.2 tutorial to soon (i even expect it for 6.2 later this year then september - but thats pure speculation)

no problem if the cpu of the host system would work as baremetal for that version/model (min. 4th gen intel core with 918+ image) as with 916+ the 918+ needs a sata disk as bootdevice (instead of the usb flash drive with baremetal or the IDE mentioned above) and the disks for data also have to be defined as sata 1st sata 50mb boot vmdk (use the wrapper file for the img file that comes with 1.02a), 2nd sata 100GB, ... dsm will "ignore" the 1st 50mb disk as it cant hold any data like system or swap partition

basically there is no difference when configuring 1.03 loader and installing dsm, mod grub.cfg as in the dsm 6.1 tutorial there is also a extra.lzma for 1.03 and at the bottom of the 1st post there are some comments about differences that might later end up in the tutorial for 1.03/dsm 6.2 there is a reason why there is no "official" tutorial, its still experimental and some drivers will not work, also there is only one modell (918+) supported so every update is a migration to a new model and you might loose plugins as there is no support on the new system and you cant reinstall them best way of trying is to have a separate usb flash drive with 1.03 loader and a empty separate disk and disconnecting 1.02 loader usb and all dsm 6.1 disks, that way you can find out whats working (hardware, plugins), if you are done testing you can remove usb and the single disk, reconnect the old usb and disks and have gained the information you need what might happen if you use 918+ dsm 6.2 you can read about some of the problems here then stay off 1.03a2 and dsm 6.2, its for testing and finding out whats working and whats not, if there is a feature in 6.2/918+ you really need and its working then it might be a option (like m.2 ssd or hardware transcoding - if its working for your hardware and you get it working) i guess after there is more then 918+ image support for 1.03 it might be a time to think about having a look again

i've checked with my test system and the smart problem is also to be seen there, i never flipped open the disk in HDD/SSD, just checked if disk(s) found and created volume/raid there is also no serial number and firmware version i'd prefer to use the original kernel driver but i cant see why its crashing on load, the (older) mpt2sas seems not to be a good fit, at least its working, thats more the what i got with the 916+ image, but yes without hdd serial and smart it pretty limited one thing in the log indicates that there is something wrong dmesg mpt2sas can't register misc device [minor=221] messages 2018-07-21T00:14:34+02:00 Diskstation synostoraged: disk/disk_is_mv_soc_driver.c:71 Can't get sata chip name from patter n /sys/block/sdg/device/../../scsi_host/host*/proc_name 2018-07-21T00:14:34+02:00 Diskstation synostoraged: disk/disk_is_mv1475_driver.c:71 Can't get sata chip name from patter n /sys/block/sdg/device/../../scsi_host/host*/proc_name 2018-07-21T00:14:34+02:00 Diskstation synostoraged: SmartDataRead(108) read value /dev/sdg fail 2018-07-21T00:14:34+02:00 Diskstation synostoraged: disk/disk_temperature_get.c:104 read value /dev/sdg fail i guess from some experiments that the buisiness models use different kernel code /driver and might have additional stuff in place to handle mpt controllers the synology way its a first time for a consumer model with the 918+ as long as we dont have the recent kernel sources (synology did not publish anything, looks like they dont like the GPL way) it will be difficult to say, also 3615/3617 still use the old 3.10.105 kernel with dsm 6.2 here the crash of the mpt3sas from kernel 4.4.59

1st and 2nd partiton on the usb flash drive are FAT partitions, out of curiosity i've tried Ext2Fsd-0.69.exe and with it i can "force" windows to have a drive letter for the 1st partition but there seem to be a accessrights problem as i was not able to open it, only when i opened notepad++ as administrator i was able to edit the grub.cfg, not as comfortable as using it normal but still better as mountig it in a virtualbox linux vm

i was assuming you followed the normal howto and there is some text about the extra.lzma and how to modify grub.cfg on 1st partition and copy over the extended extra.lzma (if needed), usually you prepare/mount the img file with OSFMount and after that you write it to usb also win10 latest 2 or 3 versions can at least use the 2nd partition where the extra.lzma is located (there was one that could do both but MS changed things and atm its just the 2nd)

what version of the extra.lzma did you try i got confirmation that it worked with 4.7 so thats the min. version you will have to try there driver is bna.ko (Brocade 10G PCIe Ethernet driver", also the firmware is present (ct2fw-3.1.0.0.bin, ctfw-3.1.0.0.bin) i cant see anything in the log part you posted, search the whole log for bna or brocade (if you are using the latest extra.lzma for loader 1.02b)