IG-88

not shure what the shortest/best way is, i used linux comamnd line tools extract it and the same tools should also be able zu pack it the same way back - sorry i'm no expert at this, just trying to find just a way to do it somehow lzma to extraxt the file "extra" from the extra.lzma, the "extra" is a cpio file so cpio -idv < extra to get the files out, ... cpio --help lzma --help (or manpages or google)

then just try to change the extra.lzma file on the usb-stick to load the kerel module you also have to change the /etc/modules in that file ... DISK_MODULES="BusLogic vmw_pvscsi megaraid_mm megaraid_mbox megaraid scsi_transport_spi mptbase mptscsih mptspi mptsas mptctl ata_piix megaraid_sas mpt2sas mpt3sas" ... add your "3w-9xxx.ko" at the end

sorry i'm just guessing, its what i would try so .... looking further into the extra.lzma, there is /etc/rc.modules ... EXTRA_FIRMWARES="bnx2/bnx2-rv2p-09ax-6.0.17.fw bnx2/bnx2-rv2p-09-6.0.17.fw bnx2/bnx2-rv2p-06-6.0.15.fw tigon/tg3_tso5.bin tigon/tg3_tso.bin tigon/tg3.bin" ... so looks like the firmware file has to be mentioned in that file to work?

you are refferencing to something that was for DSM 5.2 and is from 2013 i was refferencong to something that is proven to work with DSM 6.0 from 10/2016 http://files.arcao.com/xpenology/power-off.tar.gz http://xpenology.com/forum/viewtopic.php?p=75404#p75404 and http://xpenology.com/forum/viewtopic.php?p=76373#p76373 btw. i have updated my psoting with a link to the compiled evdev.ko so with the "power-off.tar.gz" and my evdev.ko it will all be what's needed to get the button shutdown working for dsm 6.1 image for ds3615xs what is in the lastest release? what realease? there is new stuff on a daily base, any link what you refferencing to? this? viewtopic.php?p=51683#p51683 "... button.ko and evdev.ko will be added to the next boot-loader ..." (from 1/2106) trantor seems not to provide bootloaders any more, the actual stuff is build/provided by jun and in his builds there is only the button.ko, no evdev.ko (and because nobody was providing one atm i tryed it myself)

@JMRR imho you miss the point i was trying to provide the image of the ds916+ might use features that might crash the software installed if the processor does not provide the features of the Pentium N3710 used for hardware based trancoding your i5-2400s is at least 5 years old and has a completly differnt core design i'd expect that the HD Graphics 405 will be providing the addional functions (like in a skylake dasktop processor?) i dont know what is used by synology but if i would have to guess it's "Intel® Quick Sync Video" ("Intel® Quick Sync Video delivers fast conversion of video for portable media players, online sharing, and video editing and authoring.") so if we asume that, then: http://ark.intel.com/products/91830/Int ... o-2_56-GHz -> yes, its the one synology uses in ds916+ http://ark.intel.com/products/52208/Int ... o-3_30-GHz -> no https://ark.intel.com/products/95594/In ... to-2_3-GHz -> yes

where does the *.ko come from? did you compile it from the synology kernel sources? https://sourceforge.net/projects/dsgpl/ ... z/download if you alrady have installed dsm (and the *.ko is correct) you can copy it to \usr\lib\modules\ (where all kernel modules are located) other way might be to add it to extra.lzma (synoboot.img\image.img\extra.lzma\extra\usr\lib\modules\) using osfmount or modify the extra.lzma that is already on the usb-stick (2nd partition)

maybe it was aboute the fact that newer linux systems use a disk id instead of /dev/sdX when adressing a disk lets say in fstab so when the port changes the disk is still used correct while adressed/identifyed by the global unique id - means even if you change the position of the drives in your raid array the raid still should work afterwards

just a few psost before viewtopic.php?p=98874#p98874 looks like the 1.02a/a2 dont have all the firmware files inside you could try to download the fw files somewhere else (http://http.debian.net/debian/pool/non- ... y.1.tar.gz) and put them into the extra.lzma (\usr\lib\firmware\bnx2\) which is part of the image.img inside the synoboot.img (unsing OSFMount)

Yeah that is what I was using, doesn't seem to boot properly though. Using the same settings on a Ubuntu live disc results in a network connection but not with the v1.02a2 img weird. Will start fresh and see how I go. i have 1.02a running 3617xs in virtual box,using the intel pro 1000/mt desktop (bridge mode with the hardware nic) and i gave the nic the mac that was used inside the synoboot.vmdk image (on IDE controller) - try to set the mac propperly so synoimage and vm nic using the same mac? for the disks used in dsm6.1 the vm has to use scsi controller/disks

afaik the SN is only for the dynamic dns function (quick connect), the image will boot with the serial already in jun's loader does have the J3455M the same feature (instructions) as the processor in the DS916+ (to use hardware trancoding)? you will have to check (read) for yourself the DS916+ comes with a Pentium N3710 quad-core ...

i would not count on the fact that every board producer uses the same layout to wire a multilane connector (4 ports), same goes for the producer of cables (mutlilane sas to 4x sata) so the numbering on the labels attached to the cable might not be right to what is shown, lets say in the bios (sas0, P1, P2, ...) you might fix the lables on the cable if you like so they are corrsponding to the numbering shown in bios not shure what you mean with "logical" but the portlayout you see in bios should be the same as the one you see in the store manager in dsm6.1, at least for one controller there may be differences in sequence of the controller depending on bios settings (bootable or not) or in which slot a controller is pluged in but as long as you only change the disk from one port to another you shoud see the disk "jumping" the same to a different port in the storage manager (if you have a real/original synology box the slots have numbers and will be equivalent to what you see in storage manager - disk 5 in dsm storage manager is not workking, replace disk in slot 5 of the hardware) if you once find out which hardware sata port is the which port in dsm storage manager (one controller will be a block of disks in a row) then you will know what disk to change when one is failing, to map this just write down the SN's auf the disks and which hardware port they are connected and then have a look at dsm storage manager, the disks will show up with serial numbers (bare metal or virtual with vt-d) in your case the 6 x onboard may also be 2+4 (or 4+2) and with the numbering of the two 4xmultilane ports there will be also two blocks of 4 drives in dsm if that is documented it will be much easier/safer if a disk is to replace

https://en.wikipedia.org/wiki/Next_Unit_of_Computing J3455M Celeron is Apollo Lake, thats newer then broadwell or braswell so there are hopefully no missing cpu instructins when using a image based on newer hardware like the 3617xs ds916+ image was choosen because it suppors the cpu insructions for hardware based transcoding

why should 3617xs be a problem? its also 6.1 an there is kernel source availible https://sourceforge.net/projects/dsgpl/ ... z/download you would just compile the evdev.ko in the same way as for 3615xs and thats it

there is no "calculation" its just the number of ports of the controllers in a system written in a row the mentioned bord has 6 ports provided by the h97 chipset and the naming an the block diagram from the manual give no indication the there is a difference between them (sata0-5) so the number for sataportmap would not be 6 but as i understand it its only needed if there are problems and more then one controller, if you put an additional 4 port controller into the system it would be 64 (6 on the first, 4 on the second) it is still odd that 2 disks where missing, according to what you wrote there was just one disk on a different controller, all other disks (8) where on one lsi controller and should habe been found you shoud test it with a clean installation on one disk and the second disk on different ports to proof the layout and that all ports are working

yes it does, for 6.0 it seems longer known to work (->Arcao), 6.1 tested yesterday viewtopic.php?f=2&t=20216&p=98854#p98854

> Müsst auf dem Laptop nicht irgendwie die DSM oberfläche erscheinen ?? nein, die dsm oberfläche ist über das netzwerk im browser zu benutzen, lokal gibts da nur eine linux (text) console mach mal ein paar angaben welchen bootloader und welches DSM du verwendest wenn du die weboberfläche des dsm berets hast dann ist ja etwas auf disk (dsm *.pat file installiert), die dsm schreibt sich nicht auf den usb stick, der existiert als reines boot medium, das eigentliche dsm istalliert sich auf der/den festplatte(n), da wird partitioniert und am ende gibt es auf jeder platte eine dsm partition (~2,5GB?, sozusagen raid1 auf alle verwendeten platten) und der rest bleibt als unbenutzte/leere platte die man im speichermanager dann sehen sollte (und da werden dann diskgruppen/volumes konfiguriert auf denen man daten ablegt) > Wenn ich den Laptop.mit DSM vom PC aus neu Starte Funktioniert es auch allerdings steht dann ; > DHCP.../ > Und es lädt .... > Doch dann kommt etwas von "No Boot Device Detected" obwohl der Stick drann hängt. klingt so als wenn bei einem reboot der usb stick nicht erkannt wird bzw. etwas anderes sich vordrängelt? boot reihenfolge im bios überprüfen und evtl. alles außer dem usb stick deaktivieren

hi, du machst hier ziemlich wenige angaben, ist schwer da etwas zu sagen baremetal oder virtuell install? (was für hardware bzw. welche virt. umgebung), bei ser. port würde man ja baremetal annehmen aber virt. hat theoretisch auch serielle ports DSM 6.0.2 (jun 1.0.1) oder 6.1 (jun 1.02a) das was du da bei angepasst aufzählst passt nicht zu den howto's hier im board, da wird nur die grub.cfg angepasst (im wesentlichen vid/pid des usb sticks von dem gebootet wird und mac adresse der netzwerkkarte die für DSM verwendet wird) was bei der installtion passiert bzw. was nicht (syno assistant findet nichts? was für eine netzwerkkarte ist benutzt und wird die auch unterstützt?) passiert bzw. was nicht

hi, the board seem to have 2+4 intel sata + 8 lsi sata/sas you could try to set the SataPortMap=1 (1 is default on usb stick, grub.cfg) to 428 or 248 (depending on how the 2+4 sata intel onboard satas are found) but i guess depending on boot configuration it might also be the 8 lsi's are first? (would explain that the 8 lsi disk are found an the 2 on intel are "out" of the counting for internal ports to not risk further raid problems i would suggest to disconnect all raid drives, take a empty hdd and make an new install to this drive (with the already working 6.0.2 usb stick), with that and a second disk to connect for testing you might be able to find what SataPortMap config is working, just connct the 2nd drive on one of the controllers (intel 4x, intel 2x, lsi 8x) and see what dsm finds you also should document witch controller is bootable and what the boot order is i have a asrock c2550d4i and it has 2+4 intel and 2+4 marvel ports (4224 when it comes to the mapping in dsm, but i have never try'd to make the marvel bootable or change the boot order, just default and disabled boot for additional controllers) PS: its also helpful to have a complete map ready with hdd serial numbers of drives and to witch port (hardware) they are connected, if that is compared to what dsm shows (sn's are sown) its likely to see the pattern of the mapping

hi, just to give some feedback got baremetal install working on Asrock C2550D4I (all 12 SATA ports working, just short testing) with jun 1.02a and 3615, the 3617 was not installing at all (found with syno assistant but failed while uploading the image - missing network support in 3617 for the intel onboard nic? atm no clue whats going on when uploading a pat image when booting from usb, did the vid/pid and mac thing right for shure) plan to migrate to 6.1 later after complete backup, raid array is ext4 and i want it to be btrfs so i will have to make a clean new install with all disks empty and restore files after install

hi, i got button shutdown working for 6.1 (u2, loader jun 3615xs 1.02a) with the scripts working for 6.0 (http://xpenology.com/forum/viewtopic.php?p=76373#p76373), just had to remove the *.ko files (different kernel for 6.1), the new button.ko is provided by jun's loader and the missing evdev.ko i build from the kernel files (https://sourceforge.net/projects/dsgpl/ ... z/download) in a chroot (plugin from synocommunity.com) on the DSM 6.1 (as discribed by Arcao for 6.0) barematel install on asrock C2550D4I (the 3615 is key, 3617 did not work - no install at all possible) edit: uploaded the evdev.ko for DSM_DS3615xs_15047 (DSM 6.1 bromolow) http://s000.tinyupload.com/?file_id=937 ... 8913451187