IG-88

good arguments to stay with dsm 6.1.x ? 1.03a2 is still alpha and for testing, its for people that like to waste time and try to find (impossible) solutions for problems that do not exist for other people i'm having "fun" myself with the drivers for some nic's and mpt2/3sas i'd suggest to search the log for anything unusual and have a look with "top" if anything is interfering

the 918+ is limited to 2 lan ports without mod there is e1000e and igb so two different drivers, if there is a pcie nic try to remove it, if teh board has 2 different lan chips the try to disable one, test switch to disable the other and test again i've tested both drivers on a skylake system also try your usb flash drive and a single empty disk with a fresh 918+ install, maybe some kind of upgrade problem (bonding and maybe mtu size othe then 1500 are involved?)

i've tried the mpt3sas (includes the mpt2sas code) kernel driver from synologys 4.4.59, no luck just crashes on boot, the mpt2sas i provide is from the latest external source so i dont see what i could do if it does not work on your system, i does on mine so its not a general problem with the driver the ahci is part of the synology kernel, so kind of nothing i can do here, only thing might be some tweaks to the start options in grub? https://xpenology.com/forum/topic/12859-driver-extension-jun-103a2dsm62x-for-ds918/?do=findComment&comment=92759 the chip itself is at least 4 years old so not much to expect, if you get 900mbps then its not to bad maybe i will include the latest 8.045 in the next release

on 1st try the kernel drivers for mpt2sas/mpt3sas did not work so i switched to the v20 kernel source of mpt2sas i already had from loader using with 1.02, 918+ uses kernel 4.4 and the release date of the driver is before the release date of the kernel so there might be hidden problems, my plan was to check an this if problems come up and try the kernel drivers again i've tested with a 9211-8i on a skylake with Q150 chipset in a pcie 16x slot and i was able to make one connected disk part of a raid1 what hardware did you try and did the pcie slot had a least 8x pcie lanes (have not tried what will happen when its used in a slot with less then 8 lanes), your 2 x sata onboard implies a smaller hardware with low count of pcie slots and lanes what network hardware are you using? thats ahci i guess but how do you test when there is no network access, there seems to be some parts missing on what you use and how you test just as a guess, check the cache settings of the disks, if write back cache of disks are not active then there will be less performance, synology has settings inside the disk manager for this but i dont know what exactly is done when its not working

get yourself stuff like debugger, IDA, ... reverse engineer jun's "patch" (its part of the loader) and redo it for the other systems (kernels, the loader is pretty much the same, the hack is specific for every platform aka kernel) i guess the 918+ is just the test to see if all protections are gone and the patch is 100% working, investing work on two other platforms and having to redo it because something was missing in the 1st try is not much fun i guess, if everything with 918+ goes well for a while i expect jun will release more just give it a while i guess its like solving a problem (doing one/1st platform patch) is fun and maintaining stuff (doing more platforms) is work, to make it useful for many people you will have to do both but you want to minimize the 2nd part, i think (just logic) jun will at least do a patch for 3617 later to see if potential problems are found, if 3617 can handle all hardware like 3615 it makes not much sense to maintain 3615 further, the update support of synology will end earlier for 3615, but i don't know anything about how jun's point is there so its pure speculation based on things i've catched here in the forum so be patient or start hacking (an if you are on hacking, just have a look for the sata kernel mod's synology did so we can have sata_*.ko/pata_*.ko drivers)

if you just have a basic disk in dsm then your way without backup/restore to a new bigger virtual disk would be going over to JBOD mode where disks are just chained (no redundancy like the basic disk), the equivalent to that on a normal linux might be lvm volume with 2 (different) disks, after "glueing" the partitions together as one the file system can be extended to use the newly added space (dsm does that as JBOD mode)

ich wüßte da noch eine abkürzung, dabei gehen nicht mal alle settings drauf da "migriert" wird wenn man eine 3615 hat die auf dsm 6.2 gekommen ist kann man einen 1.02b boot stick für 3617 erzeugen, die 3617 installieren (ist ja kein downgrade da man auf eine andere dsm version migriert, da spilet die installierte versionsnummer keine rolle), nach der installtion erzeugt man den stick für 3615 neu vom 1.02 image file (der alte stick hat ja bereits beim update den neuen kernel bekommen, weshalb man auch immer beim booten im synology assistant die neue version angezeigt bekomme), im synology assistant sollte jetzt wieder die alte (erste) 6.1 versionsnummer zu sehen sein, dsm 6.1.x (kann gleich 6.1.7 sein) *.pat file für 3615 installiren und das war's beim migrieren wird nur die system partiton neu geschrieben und dabei werden soweit möglich auch die settings übernommen, die daten partiton mi den eigenen daten wird dabei nicht gelöscht, da 3615 und 3617 fast gleich sind wird da so zielich alles an einstellungen mit übernommen wichtig ist das man immer wenn man den usb stick vom image file neu schreibt man dann auch vid/pid anpassen muss damit das geschribene image zum usb stick passt hat auch den vorteil man muss nichts rumstecken oder das gehäuse öffnen, nur 2 mal den usb stick neu schreben und an die box stecken

just to show you how you can estimate it Xeon E3-1225 v5 -> Skylake, reported to work (4th gen intel core aka Haswell seems min.) i219 -> igb.ko is igb.ko part of the loader? yes it is already in dsm (rd.gz) itself and so its part of juns loader 1.03a2 Storage? -> specs of hardware from dell or 3rd party site -> 4 x AHCI -> ahci is nativ part of DSM/synology kernel ----> processor, nic and storage ok, so it should work i also added your cpu to the supported list, if you successful tried video hardware acceleration transcoding (your cpu has quick sync video and might support it) write about it here,

why not using a recent desktop you already have with win7 or win10 and virtualbox vm's of DSM i've tested this 1.02b 3615/3617/916+, 103a2 918+ on a skylake and ryzen 7 2700x

in theory to 5 and then 6.1 but even if this works i think you come from 32bit and the 6.x is 64bit you will loose all plugins and some settings on vm its much easier to install a new vm and implemt it parallell to your old dsm 6.1 fresh yes but you should also create new disks as the new default filesystem is btrfs (and still ext4 is offered) if you want a in place upgrade yes but only a very few people will recommand this it shlould but on 4.2 it might have bin ext3, i guess that will still work but an new creation its not offered anymore i'd recommand new vm and new volume with new btrfs fiel system, copy data from old or do a beckup and restore to new system

that's a 32bit single core cpu, imho its not gong to work with dsm 6.1, all three images are 64bit and i think thats the same for dsm 6.0 maybe you can try this instead of dsm https://sourceforge.net/projects/openmediavault/files/3.0/openmediavault_3.0.36-i386.iso/download imho no

same as @modboxx hardware, just ahci (driver from synology dsm) and realtek 8111 (r8168.ko, v8.044.02) nothing from my extra.lzma used, just what synology and jun already provided on write speed you should be more specific, is it about network speed or is the ahci/disk speed limited realtek nic implementations are know to be "difficult" but its not a chip thing, i have pcie card with 8168e chip and tests with dsm (mainly 6.1) i have seen speeds of 112MB/s (SMB/CIFS), same as the intel nic did latest driver version by realtek is 8.045 from 9/2017 so its synologys and jun's are pretty close to recent, not much gain to expect from the latest driver your board has a pcie 1x slot so you can test a different nic if you want

~12 up ~23h ago i wrote about the new extra.lzma and 9211 is mentioned and i tested it myself with a it-mode flashed 9211-8i jun's extra.lzma did not contain mpt2sas/mpt3sas so if its just the lsi that is missing it should work with my extended extra.lzma, but as long as you are not keen to use hardware trancoding i dont see any gain, even more risks with the new loader and 918+ image

thats not much information and also implying you just exchanged the usb hardware and keeping the loader (version) and with this the former dsm version but my guess is you try a newer loader and dsm version like .loader 1.02b / dsm 6.1.x, the type seem to be 3615 from your screenshot the message indicates no disks found so, no controller (no driver) or no disks (like because its a raid controller and you cant expose single disks as needed, IT-mode like in "initiator target mode") the driver in question (from your ubuntu log) would be "mptsas.ko" and thats part of juns loader so is not "no controller found" if its a raid controller try to create a raid (one or two disks like raid0 or 1) and see if it finds disk(s) then also your server has ilo and you can use that to have a look on the serial port over network, dsm 6.x does not give much information on the screen, ist using the serial port as console, so if youn want to see whats going on user ilo's serial over network (search the forum for this) ps: sas1068 has a hardware limit about disk size so no disk bigger 2TB if you plan to extend with this controller

your controller is not in my list as supported your live ubuntu presumably using a newer kernel/driver must be a newer type/chip that was not supported by kerne 3.10.102? they seem to offer recent driver source so i can try to compile it https://storage.microsemi.com/en-us/support/raid/sas_raid/sas-71605e/ https://storage.microsemi.com/en-us/support/raid/sas_raid/asr-81605z/ it something useful comes up from this i will write about in in the additional compiled modules section

only if something does not work at the beginning its good to know on what systemboard and adtional cards it was used to see whats working also just for a bare Asrock J3445M with 2x ahci storage and a realtek onboard nic you dont need this extra,lzma, the driver for this are even inside the original dsm 6.2 file

i have uploaded the first version of my extra.lzma for 1.03a2/DSM 6.2 if anyone want to give it a try in contrary to 916+ i got mpt2sas working so the often used lsi sas 9211 will work, also as its a new kernel we might give the hp smart array controllers a new try, the driver from kernel 4.4.59 is included for testing, if that does not work we can still try the driver from external source defiantly not working (as in kernel 3.10.102 - its a synology mod thing) are all sata_* and pata_* drivers

as we are in the english section we have to stay by that language with loader 1.02a jun provided a vmdk file thats like a wrapper for the 50 MB image file you can use this vmdk (text) file and the synoboot.img (1.02b or 1.03a2) to integrate the image file as disk into virtual box that way you can still use the usual tools like osfmount to modify the virtual disk used for virtual box (like changing grub.cfg or exchange extra.lzma)

This version is alpha, as the loader 1.03a2 is, so DO NOT use this on a "production" systems and be prepared to loose the data you have on the "test-system" (I recommend you have a very recent backup) !!!all sata_* and pata_* drivers are still not working because of the kernel source changes Synology did (same as in kernel 3.10.102 used and modded by Synology), as long as no one writes a patch for this I can use to fix this, it will stay this way !!! HP SmartArray Controller driver (hpsa.ko) is back in for testing (for now the one that comes with kernel 4.4.59), my P400 i had access to was to old for this driver so please try to test, newer controller even seem to have a IT mode so if present try it with this mode To check if the hardware is supported with a driver, use a running DSM installation or boot a live Ubuntu from usb. In console type the following: lspci -k | grep 'Kernel driver' That command will list the used kernel modules for the hardware. Modules for network and storage will be listed. Compare that with the list of modules that comes with the extra.lzma ramdisks provided below. extra.lzma for DS918+ v0.2 (still online for testing against new 0.4) http://www.filedropper.com/extra918plusv02 SHA256: 793B844F2649B81D9B3096CD458E169DB254989FC8B2519341FA464D517D00C2 Mirrors by @rcached https://dailyuploads.net/frna1z7ba8bo https://usersdrive.com/t71xwwy74un9.html https://www23.zippyshare.com/v/NSpdJ5Xc/file.html extra.lzma for DS918+ v0.3 http://www.filedropper.com/extra918plusv03 SHA256: 452ACB232607E15CFC5D88EFBF03800EA034383FC0584A2CDB2180CEA49C0660 Mirrors by @rcached https://dailyuploads.net/63jy78stbfq1 https://usersdrive.com/3svgjiyx13im.html https://www78.zippyshare.com/v/7ZYKeVbl/file.html extra.lzma for DS918+ v0.4 http://www.filedropper.com/extra918plusv04 SHA256: D7C5150BA0EF1653B3F3B688793B10E7503589E75CBA0E1D6D054388C14606DE Mirrors by @rcached https://dailyuploads.net/ud7lq6m2bto9 https://usersdrive.com/l34neyil9xcf.html https://www74.zippyshare.com/v/45PDyF86/file.html extra.lzma for DS918+ v0.5_test http://www.filedropper.com/extra918plusv05test SHA256: 765095FC170CB1A23371E6BD7A593F9FD8CA3E0B2D86445C25DA72420FCBBF04 Mirrors by @rcached https://dailyuploads.net/i1ctdis7m3hf https://usersdrive.com/4dfe3ks2tk9y.html https://www2.zippyshare.com/v/8xkz80dK/file.html SHA256: 765095FC170CB1A23371E6BD7A593F9FD8CA3E0B2D86445C25DA72420FCBBF04 Note: the driver list below with related hardware is based on Trantor/Quicknick's older list for 5.2 and 6.0 so thanks goes to them and all others who helped to build the older list for 5.2 and 6.0. Caption (…) Not compiled for now, ask for it […] Either confirmed by users that it is not working OR not supported for DSM 6.2 (kernel 4.4.59) OR does not load at all. Driver was compiled/tested but removed - if someone finds a way to get it working in some way, tell us here and it will be added  If you wish to make a driver request do so in the topic linked below: If you wish to discuss problems about modules in these ramdisks then drop a post below. Please be specific and provide relevant information such as: ramdisk version, loader model, motherboard and any additional storage/networks cards present in the system. Other things good to know about DS918+ image and loader 1.03a2: 1. DS918+ image comes with a limit for 2 internal network ports in synoinfo.conf (the original hardware has two ports and does not have a pcie slot for extending) jun's loader does not changes this (now?), so if you plan to user more then 2 ports you will have to manually change the synoinfo.conf (3615 and 3617 are bigger business models and have options for additional network card, the default from synology is 8 on this systems) further information can be read here: 2. hardware transcoding seems to crash the system when a monitor is attached, unplug monitor, reboot try again to see if there is decoding hardware present you can try this (there should be something in like card0) ls -al /dev/dri/ when using plex you should download the 64bit version from website, with plex pass active (hardware-accelerated streaming is a premium feature and requires an payed active Plex Pass subscription), Settings > Server > Transcoder, Turn on Show Advanced, Turn on Use hardware acceleration when available 3. if your screen goes blank during boot and the system does not show up in network the kernel might have crashed on boot while loading i915.ko driver (CPU internal GPU) look in your bios for a legacy mode option, disable it and try again (boot process and crash can be seen when using a "good old" serial cable and a console program like putty on a 2nd computer - that's rarely done but if you want to know for sure, that's your way to find out, DSM does not have a typical boot screen that shows log info and login prompt after boot, that's redirected to the serial port)

yes and yes dont user ide disk for 50mb boot image, use sata 1st disk as boot media (instead the usb on bermatel), 2nd sata disk will be the "normal" disk you are installing to (like 100GB thin disk) nic intel 1000PRO/MT, the mac has to set in the nic features in the vm to the same as on the boot media (grub.cfg)

not a general problem, tried a fresh install from a usb3 port on my skylake test system, no problems, clean install but should not be ignored when having error 13 problems, did see error 13 problems (vid/pid and pat file ok) on my C2550 system when testing 918+, might be something different but at least its not always wrong vid/pid

i will add a comment about this one in m ylist strange as J3455 is exactly the cpu that comes with the 918+

don't hold back about the nic, if we know what vendor/type (if its onboard neme the board) someone might be able to make a estimate (at least i will) i5-2500k will be to old, 4th gen intel core seems to be the min. for 918+ - did you actually tried it? if yes i will add it so the not working section about the i5-4570, we had at least 3 different reports (45xx/46xx/47xx) for Haswell to work

also another warning about bios settings for 918+ image i have a hp prodesk 600 g2 for testing (skylake) and the booting was erratic and resulted often in kernel crashes after swtichg off the legacy option in bios (no legacy, no secure boot) the system comes up stable every time when booting the crash was related to 915i driver so i guess i've never seen this before as when testing 3615/3617 it does not matter and with 916+ testing i guess the i915 was not used when loading as it did not fit the (newer) hardware

just a small warning for people trying the 918+ with more then 2 NIC's (more precisely more then 2 network ports) (same goes for 916+ i guess) i had some fun with that as non working drivers and this "new" limit interfered confused me a little (driver working and no ethX) the limit is there for years but not interfered when using the bigger business models that can have additional nic's 918+ has maxlanports="2" in synoinfo.conf (in /etc AND /etc.defaults) the standard for the "normal" 3615/3617 is "8" to get more then two ports working you need to change this or you will and up with only eth0 and eth1 you can change maxlanport to 8 maxlanport="8" and also should add values for ethX_mtu (if i remember correctly they where already up to eth3 - i guess - as there are 2 wifi usb allowed as addition to the two internal) eth4_mtu="1500" eth5_mtu="1500" eth6_mtu="1500" eth7_mtu="1500" you might add wol options but imho not really needed eth2_wol_options="d" eth3_wol_options="d" eth4_wol_options="d" eth5_wol_options="d" eth6_wol_options="d" eth7_wol_options="d" hopefully after sorting this out the extra.lzma will be tested soon