IG-88 Posted April 15, 2020 Author Share #576 Posted April 15, 2020 9 hours ago, mysy said: Yes, I get a Lenovo 430-8i HBA card, it was SAS3408 chipset, aka LSI 9400-8i , IT mode firmware. spec: 12Gb SAS HBA card. I use it with Esxi 6.7 and Win 2016 with broadcom driver is working fine connect 8 HDD to 2 disk cage. Just remind, SAS3408 use driver name: ItSas35 sure? SAS3408 is dev id 00af https://pci-ids.ucw.cz/read/PC/1000 and that supported with the driver in the 0.5 extra.lzma for dsm 6.2.2 3617, mpt3sas.ko (not with 3615 or 918+) only sas38xx/39xx is not supported (at least when it comes to the device list from the driver Spoiler pci:v00001000d00000064 pci:v00001000d00000065 pci:v00001000d0000006E pci:v00001000d00000072 pci:v00001000d00000074 pci:v00001000d00000076 pci:v00001000d00000077 pci:v00001000d0000007E pci:v00001000d00000080 pci:v00001000d00000081 pci:v00001000d00000082 pci:v00001000d00000083 pci:v00001000d00000084 pci:v00001000d00000085 pci:v00001000d00000086 pci:v00001000d00000087 pci:v00001000d00000090 pci:v00001000d00000091 pci:v00001000d00000094 pci:v00001000d00000095 pci:v00001000d00000096 pci:v00001000d00000097 pci:v00001000d000000AA pci:v00001000d000000AB pci:v00001000d000000AC pci:v00001000d000000AD pci:v00001000d000000AE pci:v00001000d000000AF pci:v00001000d000000C0 pci:v00001000d000000C1 pci:v00001000d000000C2 pci:v00001000d000000C3 pci:v00001000d000000C4 pci:v00001000d000000C5 pci:v00001000d000000C6 pci:v00001000d000000C7 pci:v00001000d000000C8 pci:v00001000d000000C9 pci:v00001000d000000D0 pci:v00001000d000000D1 pci:v00001000d000000D2 pci:v00001000d000002B0 Quote Link to comment Share on other sites More sharing options...
IG-88 Posted April 15, 2020 Author Share #577 Posted April 15, 2020 3 hours ago, mosaati said: BTW, you have to increase the amount of drives by editing a config file post installing DSM, If that is what you are referring to not detecting your drives. I had 16 drives and it was limited I think to 8 and had to edit the file for DSM to see and pick up the remaining. the limit is the default from synology and its 12 drives for 3615/17 Quote Link to comment Share on other sites More sharing options...
smartire Posted April 15, 2020 Share #578 Posted April 15, 2020 56 minutes ago, IG-88 said: the limit is the default from synology and its 12 drives for 3615/17 Thank you IG-88 for the info, see I've installed successfully DS3617xs with DSM_DS3617xs_23739.pat no extra. I'm kind =d happy but always room for improvement. I afraid to update to 24922 and break it.. it not freezing but only detect 3 drives. # service limitations maxaccounts="16000" maxgroups="512" maxshares="512" maxdisks="12" maxprinters="2" maxlogsize="64" Quote Link to comment Share on other sites More sharing options...
IG-88 Posted April 15, 2020 Author Share #579 Posted April 15, 2020 37 minutes ago, smartire said: Thank you IG-88 for the info, see I've installed successfully DS3617xs with DSM_DS3617xs_23739.pat no extra. I'm kind =d happy but always room for improvement. I afraid to update to 24922 and break it.. it not freezing but only detect 3 drives. for ??? reasons most of the drives are above the 12 drive limit and wont be shown (might be visible as external drives) what controllers are in this system (onboard and as pcie card) any sas extenders? how are the disks connected (what controller) please do some information gathering lspci -k ls /dev/sd* quick and easy might be to raise the allowed disk amount to 24 but that will be gone after a bigger update (full ~250MB pat file) better to locate the problem and solve it otherwise, so so funny if you raid breaks after such a update from above it should be solved by tweaking the the loader, bios or hardware itself Quote Link to comment Share on other sites More sharing options...
smartire Posted April 15, 2020 Share #580 Posted April 15, 2020 (edited) 47 minutes ago, IG-88 said: for ??? reasons most of the drives are above the 12 drive limit and wont be shown (might be visible as external drives) what controllers are in this system (onboard and as pcie card) any sas extenders? how are the disks connected (what controller) please do some information gathering lspci -k ls /dev/sd* quick and easy might be to raise the allowed disk amount to 24 but that will be gone after a bigger update (full ~250MB pat file) better to locate the problem and solve it otherwise, so so funny if you raid breaks after such a update from above it should be solved by tweaking the the loader, bios or hardware itself I'm following you procedure Case 1: update from 6.2.0 to 6.2.2 to upgrade, it's rebooting right now. but it still messing with BIOS time DS3617_results.txt Edited April 15, 2020 by smartire It was empity Quote Link to comment Share on other sites More sharing options...
htvsm Posted April 15, 2020 Share #581 Posted April 15, 2020 12 hours ago, Arcao said: Note: Marvel 88SE9215 also exists in M.2 M Key 2260 version: https://www.aliexpress.com/item/32814284199.html?spm=a2g0o.productlist.0.0.25b53c0ds9S1kJ&algo_pvid=e52cd22c-0354-472f-9c68-076f7c750c23&algo_expid=e52cd22c-0354-472f-9c68-076f7c750c23-6&btsid=0be3743615869391392268670eb250&ws_ab_test=searchweb0_0,searchweb201602_,searchweb201603_ Or you can buy M.2 M key to PCIe 1x riser card similar to this: https://www.aliexpress.com/item/4000825543839.html?spm=a2g0o.productlist.0.0.377d17088YtFCl&algo_pvid=89c570e2-7219-41e7-8dcf-4ba32f7fdaa0&algo_expid=89c570e2-7219-41e7-8dcf-4ba32f7fdaa0-24&btsid=0ab6fa7b15869398135633409e5814&ws_ab_test=searchweb0_0,searchweb201602_,searchweb201603_ Hi arcao, thanks for your suggustion. Already put on my wishlist. I considered the M.2 pcie type . Thank you bro.👍 3 hours ago, IG-88 said: i know this one, with all this corona trouble my delivery if way off schedule pcie is usually backward compatible, i'd expect it to work in a pcie 2.0 slot Nice, yeah.. also in my country all shipment from outside is freeze currently. Maybe i will purchase until my country movement control order is lift off. Hopefully all around the world is back to normal asap, i already stayed at home about 4 weeks already.😐 Quote Link to comment Share on other sites More sharing options...
smartire Posted April 15, 2020 Share #582 Posted April 15, 2020 The upgrade was successful here the Controller model: PERC H730 Mini RAID Controller 1 Quote Link to comment Share on other sites More sharing options...
smartire Posted April 15, 2020 Share #583 Posted April 15, 2020 13 minutes ago, smartire said: I'm following you procedure Case 1: update from 6.2.0 to 6.2.2 to upgrade, it's rebooting right now. but it still messing with BIOS time DS3617_results.txt 0 B · 0 downloads The upgrade was successful here the Controller model: PERC H730 Mini RAID Controller 1 Quote Link to comment Share on other sites More sharing options...
IG-88 Posted April 15, 2020 Author Share #584 Posted April 15, 2020 1 hour ago, smartire said: I'm following you procedure Case 1: update from 6.2.0 to 6.2.2 to upgrade, it's rebooting right now. but it still messing with BIOS time DS3617_results.txt 7.77 kB · 1 download from lspi its the C610 chipset ahci with 4+2 ports and a sas3108 controller (8 ports) 0000:00:11.4 Class 0104: Device 8086:8d66 (rev 05) Subsystem: Device 1028:0601 Kernel driver in use: ahci -> C610/X99 series chipset SATA Controller [RAID mode] 0000:00:1f.2 Class 0104: Device 8086:8d06 (rev 05) Subsystem: Device 1028:0601 Kernel driver in use: ahci -> C610/X99 series chipset SATA Controller [RAID mode] 0000:02:00.0 Class 0104: Device 1000:005d (rev 02) Subsystem: Device 1028:1f49 Kernel driver in use: megaraid_sas -> MegaRAID SAS-3 3108 [Invader] aka PERC H730 Mini RAID Controller 1 the drive list might indicate you have used 3 x sata and 8 sas ports /dev/sdi - /dev/sdk /dev/sdl /dev/sdm /dev/sdn /dev/sdo /dev/sdp /dev/sdq /dev/sdr - - /dev/sdu /dev/sdv strange is the "missing" sda to sdh, thats 8 ports so the layout for dsm looks like this 8 ??? 2 ahci 8 sas3108 4 ahci these 8 ports shift most of the drives into the >12 area, no idea where they come from nut even without these leading 8 ports you would be above 12 ports with 2 +8+4 = 14 and you would be needing to find a way to place the two last drives in a way that they would be /dev/sds and /dev/sdt can you send me a copy of /var/log/dmesg ? there might be a clue to see why 8 ports are skipped did you try loader 1.04b again? if that would be used then you could change the drive count in the patch og the loader to 24 and make the drive number persistent for updates if you want to go the way of modding the /etc.defaults/synoinfo.conf then you would have to change it like this (and reboot after changing it) original: maxdisks="12" esataportcfg="0xff000" internalportcfg="0xfff" usbportcfg="0x300000" new: maxdisks="24" internalportcfg="0xffffff" esataportcfg="0x0" usbportcfg="0x3000000" Quote Link to comment Share on other sites More sharing options...
smartire Posted April 15, 2020 Share #585 Posted April 15, 2020 51 minutes ago, IG-88 said: from lspi its the C610 chipset ahci with 4+2 ports and a sas3108 controller (8 ports) 0000:00:11.4 Class 0104: Device 8086:8d66 (rev 05) Subsystem: Device 1028:0601 Kernel driver in use: ahci -> C610/X99 series chipset SATA Controller [RAID mode] 0000:00:1f.2 Class 0104: Device 8086:8d06 (rev 05) Subsystem: Device 1028:0601 Kernel driver in use: ahci -> C610/X99 series chipset SATA Controller [RAID mode] 0000:02:00.0 Class 0104: Device 1000:005d (rev 02) Subsystem: Device 1028:1f49 Kernel driver in use: megaraid_sas -> MegaRAID SAS-3 3108 [Invader] aka PERC H730 Mini RAID Controller 1 the drive list might indicate you have used 3 x sata and 8 sas ports /dev/sdi - /dev/sdk /dev/sdl /dev/sdm /dev/sdn /dev/sdo /dev/sdp /dev/sdq /dev/sdr - - /dev/sdu /dev/sdv strange is the "missing" sda to sdh, thats 8 ports so the layout for dsm looks like this 8 ??? 2 ahci 8 sas3108 4 ahci these 8 ports shift most of the drives into the >12 area, no idea where they come from nut even without these leading 8 ports you would be above 12 ports with 2 +8+4 = 14 and you would be needing to find a way to place the two last drives in a way that they would be /dev/sds and /dev/sdt can you send me a copy of /var/log/dmesg ? there might be a clue to see why 8 ports are skipped did you try loader 1.04b again? if that would be used then you could change the drive count in the patch og the loader to 24 and make the drive number persistent for updates if you want to go the way of modding the /etc.defaults/synoinfo.conf then you would have to change it like this (and reboot after changing it) original: maxdisks="12" esataportcfg="0xff000" internalportcfg="0xfff" usbportcfg="0x300000" new: maxdisks="24" internalportcfg="0xffffff" esataportcfg="0x0" usbportcfg="0x3000000" b4_m_synoinfo_dmesg after_m_synoinfo_dmesg new_results.txt Quote Link to comment Share on other sites More sharing options...
smartire Posted April 16, 2020 Share #586 Posted April 16, 2020 (edited) I have changed the /etc.defaults/synoinfo.conf as you suggested please kindly see the results. I'm assuming that the loader 1.04b must be with DS918 correct? if yes it wont load it get stuck with error The file is probably corrupted. (13). Please let me know if you need anything from me and thank you again. if you want i can Switch to HBA mode or set the drives to non-raid. Edited April 16, 2020 by smartire Quote Link to comment Share on other sites More sharing options...
smartire Posted April 16, 2020 Share #587 Posted April 16, 2020 26 minutes ago, smartire said: I have changed the /etc.defaults/synoinfo.conf as you suggested please kindly see the results. I'm assuming that the loader 1.04b must be with DS918 correct? if yes it wont load it get stuck with error The file is probably corrupted. (13). Please let me know if you need anything from me and thank you again. if you want i can Switch to HBA mode or set the drives to non-raid. Update= I have disable the onboard controller where I had only the CD-ROM as caddy for SSD connected adn all the drives are there. reverting the /etc.defaults/synoinfo.conf to default. Quote Link to comment Share on other sites More sharing options...
IG-88 Posted April 16, 2020 Author Share #588 Posted April 16, 2020 13 minutes ago, smartire said: Please let me know if you need anything from me and thank you again. can you read the image from your working 3617 and the 918+ loader with "Win32DiskImager 1.0" with activ "read only allocated partitions", apload it somewhere (tinyupload is easy) and pm me the links? the conrolle config from dmesg is 4+6 ahci ports and 8 from sas, so its sda to sdr and 18 ports sdu and sdv are the usb drive and a virtual cd/floppy - so they dont count as dsm drives can you disable the 6port ahci controller in bios? 4+8 would a good combination Quote Link to comment Share on other sites More sharing options...
IG-88 Posted April 16, 2020 Author Share #589 Posted April 16, 2020 4 minutes ago, smartire said: reverting the /etc.defaults/synoinfo.conf to default. from memory i thought it would be enough to change the synoinfo.conf in /etc.defaults and then reboot to see the change in /etc/synoinfo.conf too so it you do the change for both it should work Quote Link to comment Share on other sites More sharing options...
djvas335 Posted April 16, 2020 Share #590 Posted April 16, 2020 It seems that extra.lzma for ds3615xs does not work with DSM 6.2.3-25423, I have upgraded a test microserver gen7 and it didn't come up. Quote Link to comment Share on other sites More sharing options...
IG-88 Posted April 16, 2020 Author Share #591 Posted April 16, 2020 45 minutes ago, djvas335 said: and it didn't come up. did you use a serial port as console or are you just guessing as it did not show up in network? i still have a test hardware with a 1G broadcom nic so i might be able to test it @hardware transcoding users i had a quick glance at the i915 driver from 6.2.3, it is from 22.12.2017, newer then the one in 6.2.2 (19.09.2016) but slightly older then the i915 created by jun for loader 1.04b (14.05.2018) there are firmware references to "glk_dmc_ver1_04.bin" so it should support gemini lake there are some already people that dont have transcoding support after update without digging in i think the reason will be missing firmware, 918+ original image does not contain all i915 firmware files, just the one needed for apollo lake (bxt), it also contains glk firmware that indicates that gemini lake should work ootb without any additional extra.lzma changes so people with "normal" gen4-9 cpu's should be fine with a updated extra/extra2 "syno" that would contain additional i915 firmware files for people who cant wait firmware files referenced in new 6.2.3 i915 driver: bxt_dmc_ver1_07.bin bxt_guc_ver9_29.bin bxt_huc_ver01_07_1398.bin cnl_dmc_ver1_06.bin glk_dmc_ver1_04.bin kbl_dmc_ver1_04.bin kbl_guc_ver9_39.bin kbl_huc_ver02_00_1810.bin skl_dmc_ver1_27.bin skl_guc_ver9_33.bin skl_huc_ver01_07_1398.bin firmware download from here: https://git.kernel.org/pub/scm/linux/kernel/git/firmware/linux-firmware.git/tree/i915 copy files in dsm to \usr\lib\firmware\i915\ Quote Link to comment Share on other sites More sharing options...
neo82ita Posted April 16, 2020 Share #592 Posted April 16, 2020 With the jun1.04b and extra 0.8 (all 3) I tried to put the ds918 + on the asus p8b75-m lx plus and i3 3225, but there was no way to start it. On the same configuration but with jun1.03b and extra 0.5 the ds3617 + has started correctly.Unfortunately I don't have the serial to see what's wrong, even if I think it's a LAN problem, he didn't find it on the net.Have some suggestions?Inviato dal mio iPad utilizzando Tapatalk Quote Link to comment Share on other sites More sharing options...
djvas335 Posted April 16, 2020 Share #593 Posted April 16, 2020 I assume as in this thread, all updates to DSM 6.2.3 which use extra.lzma were also unsucessful, however all other systems which do not use extra.lzma were all sucessful. Quote Link to comment Share on other sites More sharing options...
IG-88 Posted April 16, 2020 Author Share #594 Posted April 16, 2020 18 minutes ago, neo82ita said: ds918 + on the asus p8b75-m lx plus and i3 3225, cpu is to old for 918+ needs 4th gen, your cpu is 3rd gen, is documented here: https://xpenology.com/forum/topic/13333-tutorialreference-6x-loaders-and-platforms/ Quote Link to comment Share on other sites More sharing options...
neo82ita Posted April 16, 2020 Share #595 Posted April 16, 2020 cpu is to old for 918+ needs 4th gen, your cpu is 3rd gen, is documented here:https://xpenology.com/forum/topic/13333-tutorialreference-6x-loaders-and-platforms/Sorry, i hadn't noticed this.Among my various experiments, if I remember correctly, on the j1900 that I currently use with ds3617, I had put the ds918 +, assuming that it also worked on the i3.I'm trying to make a nas with hw transcoding, do you have any advice, without spending too much?Thanks again for everything Quote Link to comment Share on other sites More sharing options...
IG-88 Posted April 16, 2020 Author Share #596 Posted April 16, 2020 16 minutes ago, neo82ita said: if I remember correctly, on the j1900 that I currently use with ds3617, from reports i seems the J1900 will run with 918+, dsm 6.2.2 and the "syno" extra/extra2 with /dev/dri so hardware transcoding should work (in the limits what a J1900 can do) the J1900 gpu is ivy bridge level, thats good to see whats possible https://en.wikipedia.org/wiki/Intel_Quick_Sync_Video 21 minutes ago, neo82ita said: I'm trying to make a nas with hw transcoding, do you have any advice, without spending too much? maybe wait a few days to see if gemini lake now works with dsm 6.2.3, if thats the case there should be some reasonable priced units or did you have something in mind with more sata ports or 10G nic? Quote Link to comment Share on other sites More sharing options...
premikkoci Posted April 16, 2020 Share #597 Posted April 16, 2020 I tried to update my j4105 to 6.2.3 and unfortunately there is no /dev/dri/. I use juns loader 1.04b without any additional extra.lzma. Quote Link to comment Share on other sites More sharing options...
IG-88 Posted April 16, 2020 Author Share #598 Posted April 16, 2020 (edited) 26 minutes ago, premikkoci said: I use juns loader 1.04b without any additional extra.lzma. yes and thats the problem, juns loader contains drivers for 6.2.0 and it contains i915 drivers replacing the drivers from synology so even with 6.2.3 juns drivers will be loaded you need to delete juns driver in the update folder (rm -rf /usr/lib/modules/update/*), shut down and then put the "syno" extra/extra2 to your loader when booting the drivers will be back in update folder but without i915 drivers, so dsm's own i915 will be used (and with 6.2.3 it should be with gemini lake support) Edited April 16, 2020 by IG-88 Quote Link to comment Share on other sites More sharing options...
premikkoci Posted April 16, 2020 Share #599 Posted April 16, 2020 I started a downgrade process a few minutes before your answear. I am gonna try a clean install then. Quote Link to comment Share on other sites More sharing options...
IG-88 Posted April 16, 2020 Author Share #600 Posted April 16, 2020 just a quick warning for 6.2.3 attempts, all additional drivers from 6.2.2 will fail (crash) to load so if you update only "native" drivers from DSM will work the mentioned i915 firmwares can be loaded as expected to get all non apollolake/geminilake cpu's to work (tested with a 8th gen cpu, got /dev/dri devices) looks like there is a reason why its a complete new version 6.2.3 Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.