Jump to content
XPEnology Community

sbv3000

Moderator
  • Posts

    1,061
  • Joined

  • Last visited

  • Days Won

    10

Everything posted by sbv3000

  1. From your description you updated using a non tcrp loader to 7 then used a tcrp boot/build? And again from description you created a 7.0.x loader but installed 7.1.x? I would start afresh with a new tcrp boot stick, build with a version match to the dsm version to be installed, provided all your hardware is found you should be able to migrate and install 7.1.x and retain settings Sent from my NOTE 20 using Tapatalk
  2. i would check carefully the mobo bios and sata setup, this spec says 6+2 3gb+6gb sata ports, https://bit-tech.net/reviews/tech/motherboards/asus-p6x58d-e-review/1/ disable the firewire, audio, check sata ahci, maybe disable the 2 6gb sata (if possible) that makes the setup easier with 6 drives, blank values as Peter says or try "6"/"00" or "62"/"0006" or m shell script I've tested a similar board from the same era with similar specs with tcrp build works ok with the right bios settings.
  3. as raid 0 doesnt have redundancy you might have problems, but you could try this tutorial from synology https://kb.synology.com/tr-tr/DSM/tutorial/How_can_I_recover_data_from_my_DiskStation_using_a_PC or maybe try some third party recovery tools http://www.freeraidrecovery.com/library/raid0-recovery.aspx
  4. server mobos generally have this feature, an ip kvm, there are add on cards based on r-pi might work https://www.youtube.com/watch?v=cVWF3u-y-Zg other option is to run in esxi if your hardware supports it, you always have 'console access' so could access grub menu
  5. you can make the edits to user_config.json before the build with TCRP. If I were in your situation I would create a config/boot disk for your 'base' build then rebuild, recreating satamap and port configs for each of the disk expansion phases. this would also limit the chance of usb hdd appearing as an internal disk, but maybe a better option is to unplug it before boot
  6. ok so what I determined with my test setup is; the usb boot disk mounts (hidden) as sd25 a usb hdd mounts as sd26 but shows up as internal hdd disk 21 - as seen and never less than 21 if a disk is attached to port 21 the usb hdd mounts as sd26 but appears as internal disk 22 etc etc i'm out of ideas
  7. lol I did some more tests too (but just config settings) The default usbportcfg seems to be 0x8700, I look at two of my 'standard' boxes, ds3615-42661, I added a line to user_config.json of "usbportcfg": "0x000000", rebuilt and on checking via telnet both the etc and etc/defaults were updated to this, however, it still doesn't solve the 'mount on first free sata port' that you are seeing. I'll see if I can find an old usb hdd adapter to play with and come back to you on that.
  8. Ive not tested in the way you have, but in your situation a system with <20 drives the USB HDD mounts 'normally' but > 20 it mounts as the n+1 internal HDD? If so then I wonder if its worth testing both scenarios, and checking the dmesg output for the device being mounted and any differences in the 'mount' process. Another thought is the usb hdd adapter hardware, maybe some weird situation with buses adapters etc whereby > 20 drives cant allocate resources or read the usb hdd command set properly and so it 'defaults' to mounting as a 'fixed' internal drive?
  9. Hi guys, to add something to this topic, I put together (using a single hdd) a 24 port test system 4 on board, 4*4 and 1*4 add on cards (no further pci slots on the mobo) , set maxdisk to 24 and internalportcfg to 0xffffff in user_config.json. Booted with hdd on onboard sata 1, telnet to check synoinfo.conf (both versions) and they were 'updated' with those values, and a timestamp of the boot time. Moved hdd to port 24, booted successfully, telnet, synoinfo.conf still with 24/0xffffff and the file timestamp was the new boot time. So, looks like values in user_config.json are pushed though to the .conf files each boot, which makes sense and so no reason to think that with the right values in there for a large array wont 'stick'
  10. @-iliya- is working on the same topic, I suggested removing any disk config entries in user_config.json, see thread Sent from my BAH-W09 using Tapatalk
  11. Any Intel e1000 based nic Also, some dumb suggestions Reset bios, disable any serial/parallel/audio peripherals Try an xpe5.2 basic boot loader with spare hdd, (old but works with most hw) find mac, install dsm, telnet, get pci and module info, note for your hardware. Then try dsm6 using the same boot/dsm version shown by Peter as working. If all OK reconnect your drives and migrate Sent from my BAH-W09 using Tapatalk
  12. Understood. I haven't played with large installs with DSM7 but did with 5 and 6 and recall that there are two locations for synoinfo.conf, one is /etc and the other is /etc.defaults, possible you need to amend both to stop the default one overriding the config? Also maybe remove the various entries from user_config.json? Maybe, if you havent already tried this, set the usb value to zeros, *everywhere* I have edited user_config.json after build with the built in tcrp editor, and simply run the 'build' command, other parameters unchanged. I think its also possible to mount the usb on a linux desktop and edit directly, but I've not done that.
  13. this is a good article from another forum that describes how to convert the binary drive ports to hex values for the config, there is a similar thread here but I couldn't find it https://xpenology.club/fix-xpenology-problems-viewing-internal-hard-drives-esata-hard-drives/ also have a look here http://cloudautomation.blog/2017/12/xpenology-synology-nas-use-external-usb-hard-drives-in-a-raid-group-updated/
  14. would you happen to have a spare intel nic? try that with the realtek disabled, I've read in the past that there may be 'conflicts' with versions post install of dsm, intels 'just work', however as @Peter Suh has shown, you should be able to install with your hardware without errors.
  15. in my esxi lab setup testing dsm i added two nics (on separate vswitches but I dont think it matters) booted the system, noted the two ips and mac addresses, then used the macs in the build process, serialgen realmac> save user_config.json, edit for net_if/mac2 > build, worked ok.
  16. errors around 50% installing the pat file generally means an incorrect vid/pid setting on the boot usb, check carefully the settings. also make sure you are using the matching pat file version to your build, obvious I know but worth checking
  17. check this thread https://xpenology.com/forum/topic/21663-driver-extension-jun-103b104b-for-dsm622-for-3615xs-3617xs-918/ and also follow the process using ofsmount to edit the image for grub vid/pid etc
  18. understood, I hadn't spotted the reported config limitations with lsi controllers. might be worth a search of the forum for usbportcfg threads and hba, there is a lot of history of people trying various settings to 'map' the usb beyond the hdd range, even back to 5.2 and gnoboot, sometime the hex sequences are not 'logical' the answer might be there
  19. ds1621 probably 'worked' because of the device tree architecture that builds on the fly hardware spec. if you are 'working' and happy then job-done
  20. you might want to check the sataportmap and diskidxmap, they might be 'wrong' for your hardware setup, maybe 060816 and 00060d might be my guess, plus try making usbportcfg all zeros and the string length same as internalportcfg and esataportcfg. with a larger build like yours I'd test in stages, eg build with onboard sata, then add first addon controller, build/test, then add the third. you might need to add the lsi modules manually.
  21. Can you post the step by step tcrp command results and the saved user_config.json.? Sent from my NOTE 20 using Tapatalk
  22. I would override with 2 and 0, this should work for your setup. Strange representation of the controller but it's known that some chipsets give phantom settings. You could check the chipset spec with lspci and search the vendor info if so inclined. Adding more drives at a later date would need a rebuild of the loader. Also check the bios for enabled sata and esata, this might be adding confusion to the results Sent from my NOTE 20 using Tapatalk
  23. Your mobo spec is pretty simple so sata ahci and realtek nic should all be found by the tcrp/m shell/arpl tools. Unless you have an hba and more than 4 drives? Try a build with a spare HDD Sent from my NOTE 20 using Tapatalk
×
×
  • Create New...