simoareva

Outcome of the update: SUCCESSFUL - DSM version prior update: DSM 6.2.2-24922 Update 3 - Loader version and model: JUN'S LOADER v1.03b - DS3615xs - Using custom extra.lzma: No - Installation type: VM - ESXi 6.7u3 on HP Microserver Gen8 - Additional comments: Reboot required

I am still rebuilding the raid (did not remember it took so long for empty disks in the past), however I made some preliminary tests with a single 3.7GB file on SSD VMDK, comparing UL/DL transfer speeds of DSM6.2 (via VMXNET3) and DSM6.2.1 (via E1000e), both on SMB to a W10 machine and I say they are almost equivalent. I know E1000e is not capable of managing jumbo frames but since my main router isn't able as well i think it's a fair tradeoff. On SMB i reached an average of 30MB/s UL speed (write to vmdk) and 100MB/s DL (read from vmdk). I will test them again once the raid 5 is up (made by 4 Seagate 4TB 5900RPM - ST4000DM000) to give you the RDM performance but I think it will match the previous setup. For knowledge my H310 HBA handles 5 distinct Arrays and as environment my whole house is cabled with a Cat5E Gbit so the wall of 100MB/s from SSD read is surely provided by the cable rather than the device, while the write speed is surely due the virtualization layer.

- Outcome of the installation/update: SUCCESSFUL - DSM version prior update: Fresh Install - Loader version and model: JUN'S LOADER v1.03b - DS3615xs - Using custom extra.lzma: NO - Installation type: VM - HP Microserver GEN 8 - ESXi 6.7 + 1 VMDK on SSD + 4 RDM Drives - Additional comments: See below ======================================================================================== * Modded synoboot IMG (50MB) via osfmount, loaded as a SATA disk (0:0) <- boot from this partition (change SataPortMap=88 , DiskIdxMap=0 in grub.cfg) and set S/N & MAC accordingly (used OLD SN format) * System partition 20GB on a SATA disk (0:1) * Use NIC E1000e (VMXNET3 does not work anymore) * 4 RDM mounted as SATA on dedicated Controller (1:0, 1:1, 1:2, 1:3) (Because SCSI devices are no more recognized, so you can also delete SCSI controller) * open-vm-tools.spk 6.1_10.2.0-1

I was able to make 6.2.1-23824 run on XS3615 1.03b loader on my GEN8 on top esxi 6.7 (will update to U1 soon). I gathered the information given by several members in the thread so in the end I did the following things: Edited grub.cfg on the IMG file putting SataPortMap=88 , DiskIdxMap=0 (did not change the VID/PID or serial yet) Changed every disk from SCSI to SATA (I have 4 RDM and 1 vmdk) Changed network interface from VMXNET3 to E1000e Installed open-vm-tools 10.2.0-1 Everything works ok but I am only worried about the performance of the E1000e compared to the VMXNet in terms of transfer speed.

Are there any news on progress about the modded loader for 3615xs / 3617xs? My Gen8 is getting sad

Ok I managed it. On ESXi 6.7 i setted up legacy BIOS on the VM and it was enough to put the disk on SATA instead that on SCSI (the rest of the disks remained the same). Then I migrated from VM Version 14 from 6.1.2 to 6.2 U2 with no problems. (3615)

Bare Metal or Virtualized? What settings did you used? I tried both the images (3615 and 3617) converted in VMDK or mounted directly as .img but i am stuck at the find.synology.com with no network detected (on ESXi 6.7).

what is not clear to me is if the cpu is currently not supported, it might be with an update of the loader (maybe based on bromolow like the previous one) or are we completely cut out?

Tested on a VM on a ESXi 6.7 on a Gen8 with Xeon E3-1220 V2 and it is not working. It gets stuck on this screen (as expected) but the network interface is not recognized so i am not able to trigger the setup (and yes i selected the E1000 as network card). I think it's dependant on the instructions library missing on older CPU like someone said before.

I managed to install DSM 6.1.1 (and later Update 2) on ESXI 6.5 on my Gen8. This is the procedure i followed: - Dismounted all the RDM Drives for security reason - Created a new VM from scratch with 2 disks, the first from Jun (1.02a for ds3615xs) boot image, the second empty for the OS (20GB) - Installed the latest version and run the update to reach the latest version (DSM 6.1.1-15101 Update 2) - Hotmounted the RDM disks in the VM (that were correctly recognized but unusable) - Created the Raid Group and the volume for the OS disk (the 20GB one) in order to have the same Raid Group # for the 4 data disks - Rebooted the VM - The RDM disks were correctly mounted with no further work, their Raid Group and Volume were preserved. - Reinstalled all the packages and the backup of their configurations

Can someone confirm if compiling a new kernel with the generic scsi dvd drive support it would be possible to use the drive in a guest OS (via passthrough) on virtualbox (to burn some dvds on windows)? If yes, are there any precompiled kernels for this goal? The tutorials aren't so noob friendly nor they seems outdated for the current DSM version...

I setted up everything in virtualbox and phpvirtualbox, however I was thinking I could be able to passthrough the DVD Drive of my N36L in order to use the VM just to burn some DVD when needed. I tried directly on SATA and while I can correctly see the drive with cat /proc/scsi/scsi, the device isn't obviously mounted (due the lack of drivers? I'm a total noob in linux nor I have any idea on how to do it). The guest OS consequently isn't able to use it. Any tips on how could I fix this thing? I would prefer avoiding to put XPE on a VM just for the dvd drive on another OS... Thanks