Pengy

Just leaving notes here for a problem I faced where the network speeds in my Windows 11 VM would be much slower than from the Unraid host. Speeds would be about 300mbps on iperf3 and all over the place depending on what speed test site I used. This also affected a USB NIC I tried. I was able to achieve 15+gbps to the vm host with the default number of virtio threads after this fix. I don't know what disabled the receive window tuning though. TL;DR Set Receive Window Autotuning Level to Normal running "Set-NetTCPSetting -AutoTuningLevelLocal Normal" in an elevated powershell prompt. You can check your current settings by running Reference: https://learn.microsoft.com/en-us/windows-server/networking/technologies/network-subsystem/net-sub-performance-tuning-nics Specs: AMD Ryzen R9 5900X Asus ROG Strix B550-E Motherboard (Intel I225-V 2.5G NIC) 64 GB DDR4-3600 Nvidia GTX 1660 Ti Nvidia RTX 2070 Network: Onboard Intel I225-V 2.5G NIC @ 1G Ubiquiti Dream Machine SE AT&T Fiber 1G Symmetrical Unraid Network Configuration: VM XML: Relevant Network block: <interface type='bridge'> <mac address='52:54:00:25:50:ee'/> <source bridge='br0'/> <target dev='vnet1'/> <model type='virtio'/> <alias name='net0'/> <address type='pci' domain='0x0000' bus='0x01' slot='0x00' function='0x0'/> </interface> Data Points: Before: Speed Test Screenshots: Iperf3 Runs: 10.39.0.1 - Router 10.39.0.10 - Unraid Host After: Speed Test Screenshots: Iperf3 Runs:

I figured out the problem. I should have ran memtest as a part of my general debugging step for any kernel issue, but I didn't. One of my sticks was bad and I took it out and everything is fine now.

Jun 22 07:27:09 MikuServer kernel: ------------[ cut here ]------------ Jun 22 07:27:09 MikuServer kernel: kernel BUG at fs/buffer.c:3431! Jun 22 07:27:09 MikuServer kernel: invalid opcode: 0000 [#1] SMP PTI Jun 22 07:27:09 MikuServer kernel: Modules linked in: xfs md_mod e1000e igb ptp pps_core i2c_algo_bit x86_pkg_temp_thermal intel_powerclamp coretemp kvm_intel kvm crct10dif_pclmul crc32_pclmul crc32c_intel ghash_clmulni_intel pcbc aesni_intel aes_x86_64 crypto_simd glue_helper cryptd wmi_bmof intel_cstate intel_uncore intel_rapl_perf i2c_i801 i2c_core nvme ahci mpt3sas libahci nvme_core raid_class scsi_transport_sas wmi video thermal fan backlight button acpi_pad [last unloaded: pps_core] Jun 22 07:27:09 MikuServer kernel: CPU: 1 PID: 778 Comm: kswapd0 Not tainted 4.14.49-unRAID #1 Jun 22 07:27:09 MikuServer kernel: Hardware name: Gigabyte Technology Co., Ltd. H370 WIFI/H370N WIFI-CF, BIOS F3 02/13/2018 Jun 22 07:27:09 MikuServer kernel: task: ffff88045c02aa00 task.stack: ffffc90001a54000 Jun 22 07:27:09 MikuServer kernel: RIP: 0010:free_buffer_head+0xd/0x3f Jun 22 07:27:09 MikuServer kernel: RSP: 0018:ffffc90001a57b88 EFLAGS: 00010206 Jun 22 07:27:09 MikuServer kernel: RAX: ffef88026549fe18 RBX: ffff88026549fdd0 RCX: ffffea000d72d0c0 Jun 22 07:27:09 MikuServer kernel: RDX: ffff88026549fe18 RSI: ffffc90001a57b90 RDI: ffff88026549fdd0 Jun 22 07:27:09 MikuServer kernel: RBP: ffff880288864348 R08: ffffea000d72d0c0 R09: ffff88026549fdd0 Jun 22 07:27:09 MikuServer kernel: R10: ffff88026549fdd0 R11: ffffea000d72d0c0 R12: 0000000000000001 Jun 22 07:27:09 MikuServer kernel: R13: ffff8802888642c8 R14: ffffc90001a57ca8 R15: ffffea000d72d0e0 Jun 22 07:27:09 MikuServer kernel: FS: 0000000000000000(0000) GS:ffff88046f480000(0000) knlGS:0000000000000000 Jun 22 07:27:09 MikuServer kernel: CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 Jun 22 07:27:09 MikuServer kernel: CR2: 00001488efa9fc60 CR3: 0000000001c0a005 CR4: 00000000003606e0 Jun 22 07:27:09 MikuServer kernel: Call Trace: Jun 22 07:27:09 MikuServer kernel: try_to_free_buffers+0x9d/0xaf Jun 22 07:27:09 MikuServer kernel: xfs_vm_releasepage+0x4b/0x4e [xfs] Jun 22 07:27:09 MikuServer kernel: shrink_page_list+0x7cf/0xaa2 Jun 22 07:27:09 MikuServer kernel: shrink_inactive_list+0x268/0x3f0 Jun 22 07:27:09 MikuServer kernel: shrink_node_memcg+0x4cb/0x687 Jun 22 07:27:09 MikuServer kernel: ? shrink_node+0xce/0x29b Jun 22 07:27:09 MikuServer kernel: shrink_node+0xce/0x29b Jun 22 07:27:09 MikuServer kernel: kswapd+0x43c/0x564 Jun 22 07:27:09 MikuServer kernel: ? __switch_to+0xd4/0x2e8 Jun 22 07:27:09 MikuServer kernel: ? mem_cgroup_shrink_node+0x89/0x89 Jun 22 07:27:09 MikuServer kernel: kthread+0x111/0x119 Jun 22 07:27:09 MikuServer kernel: ? kthread_create_on_node+0x3a/0x3a Jun 22 07:27:09 MikuServer kernel: ? call_usermodehelper_exec_async+0x121/0x128 Jun 22 07:27:09 MikuServer kernel: ret_from_fork+0x35/0x40 Jun 22 07:27:09 MikuServer kernel: Code: 0f c1 15 a4 00 ed 7e 81 fa ff 0f 00 00 7e 0d 48 89 04 24 e8 5b ff ff ff 48 8b 04 245a c3 48 8b 47 48 48 8d 57 48 48 39 c2 74 02 <0f> 0b 48 89 fe 48 8b 3d f0 4e b5 00 e8 2e 06 fc ff 65 ff 0d 64 Jun 22 07:27:09 MikuServer kernel: RIP: free_buffer_head+0xd/0x3f RSP: ffffc90001a57b88 Jun 22 07:27:09 MikuServer kernel: ---[ end trace d089eb21f4d110c2 ]--- Jun 22 07:27:13 MikuServer kernel: ------------[ cut here ]------------ Jun 22 07:27:13 MikuServer kernel: kernel BUG at fs/buffer.c:3431! Jun 22 07:27:13 MikuServer kernel: invalid opcode: 0000 [#2] SMP PTI Jun 22 07:27:13 MikuServer kernel: Modules linked in: xfs md_mod e1000e igb ptp pps_core i2c_algo_bit x86_pkg_temp_thermal intel_powerclamp coretemp kvm_intel kvm crct10dif_pclmul crc32_pclmul crc32c_intel ghash_clmulni_intel pcbc aesni_intel aes_x86_64 crypto_simd glue_helper cryptd wmi_bmof intel_cstate intel_uncore intel_rapl_perf i2c_i801 i2c_core nvme ahci mpt3sas libahci nvme_core raid_class scsi_transport_sas wmi video thermal fan backlight button acpi_pad [last unloaded: pps_core] Jun 22 07:27:13 MikuServer kernel: CPU: 3 PID: 35 Comm: kworker/u8:2 Tainted: G D 4.14.49-unRAID #1 Jun 22 07:27:13 MikuServer kernel: Hardware name: Gigabyte Technology Co., Ltd. H370 WIFI/H370N WIFI-CF, BIOS F3 02/13/2018 Jun 22 07:27:13 MikuServer kernel: Workqueue: writeback wb_workfn (flush-9:1) Jun 22 07:27:13 MikuServer kernel: task: ffff88045ca79c00 task.stack: ffffc90001a14000 Jun 22 07:27:13 MikuServer kernel: RIP: 0010:free_buffer_head+0xd/0x3f Jun 22 07:27:13 MikuServer kernel: RSP: 0018:ffffc90001a17338 EFLAGS: 00010206 Jun 22 07:27:13 MikuServer kernel: RAX: ffef88026569fe18 RBX: ffff88026569fdd0 RCX: ffffea000ca5f940 Jun 22 07:27:13 MikuServer kernel: RDX: ffff88026569fe18 RSI: ffffc90001a17340 RDI: ffff88026569fdd0 Jun 22 07:27:13 MikuServer kernel: RBP: ffff8801083b5248 R08: ffffea000ca5f940 R09: ffff88026569fdd0 Jun 22 07:27:13 MikuServer kernel: R10: ffff88026569fdd0 R11: ffffea000ca5f940 R12: 0000000000000001 Jun 22 07:27:13 MikuServer kernel: R13: ffff8801083b51c8 R14: ffffc90001a17458 R15: ffffea000ca5f960 Jun 22 07:27:13 MikuServer kernel: FS: 0000000000000000(0000) GS:ffff88046f580000(0000) knlGS:0000000000000000 Jun 22 07:27:13 MikuServer kernel: CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 Jun 22 07:27:13 MikuServer kernel: CR2: 0000148fd686bc60 CR3: 0000000001c0a005 CR4: 00000000003606e0 Jun 22 07:27:13 MikuServer kernel: Call Trace: Jun 22 07:27:13 MikuServer kernel: try_to_free_buffers+0x9d/0xaf Jun 22 07:27:13 MikuServer kernel: xfs_vm_releasepage+0x4b/0x4e [xfs] Jun 22 07:27:13 MikuServer kernel: shrink_page_list+0x7cf/0xaa2 Jun 22 07:27:13 MikuServer kernel: shrink_inactive_list+0x268/0x3f0 Jun 22 07:27:13 MikuServer kernel: shrink_node_memcg+0x4cb/0x687 Jun 22 07:27:13 MikuServer kernel: ? __compaction_suitable.part.4+0x5f/0x71 Jun 22 07:27:13 MikuServer kernel: ? __compaction_suitable.part.4+0x5f/0x71 Jun 22 07:27:13 MikuServer kernel: ? compact_zone+0x4a/0x533 Jun 22 07:27:13 MikuServer kernel: ? __switch_to_asm+0x24/0x60 Jun 22 07:27:13 MikuServer kernel: ? __switch_to_asm+0x30/0x60 Jun 22 07:27:13 MikuServer kernel: ? shrink_node+0xce/0x29b Jun 22 07:27:13 MikuServer kernel: shrink_node+0xce/0x29b Jun 22 07:27:13 MikuServer kernel: do_try_to_free_pages+0x17b/0x29e Jun 22 07:27:13 MikuServer kernel: try_to_free_pages+0xac/0xb2 Jun 22 07:27:13 MikuServer kernel: __alloc_pages_nodemask+0x60d/0xb31 Jun 22 07:27:13 MikuServer kernel: ? __switch_to_asm+0x24/0x60 Jun 22 07:27:13 MikuServer kernel: ? __switch_to_asm+0x24/0x60 Jun 22 07:27:13 MikuServer kernel: ? __switch_to_asm+0x24/0x60 Jun 22 07:27:13 MikuServer kernel: ? __switch_to_asm+0x24/0x60 Jun 22 07:27:13 MikuServer kernel: ? __switch_to_asm+0x30/0x60 Jun 22 07:27:13 MikuServer kernel: ? __switch_to_asm+0x24/0x60 Jun 22 07:27:13 MikuServer kernel: ? __switch_to_asm+0x30/0x60 Jun 22 07:27:13 MikuServer kernel: ? __switch_to_asm+0x24/0x60 Jun 22 07:27:13 MikuServer kernel: ? __switch_to_asm+0x30/0x60 Jun 22 07:27:13 MikuServer kernel: new_slab+0x14a/0x3ec Jun 22 07:27:13 MikuServer kernel: ___slab_alloc+0x233/0x48d Jun 22 07:27:13 MikuServer kernel: ? mempool_alloc+0x51/0x125 Jun 22 07:27:13 MikuServer kernel: ? mempool_alloc+0x51/0x125 Jun 22 07:27:13 MikuServer kernel: ? wait_woken+0x68/0x68 Jun 22 07:27:13 MikuServer kernel: ? __slab_alloc+0x17/0x29 Jun 22 07:27:13 MikuServer kernel: __slab_alloc+0x17/0x29 Jun 22 07:27:13 MikuServer kernel: kmem_cache_alloc+0xb2/0xe8 Jun 22 07:27:13 MikuServer kernel: mempool_alloc+0x51/0x125 Jun 22 07:27:13 MikuServer kernel: ? mempool_alloc+0x51/0x125 Jun 22 07:27:13 MikuServer kernel: ? generic_make_request+0xef/0x232 Jun 22 07:27:13 MikuServer kernel: bvec_alloc+0x83/0xc2 Jun 22 07:27:13 MikuServer kernel: bio_alloc_bioset+0x106/0x185 Jun 22 07:27:13 MikuServer kernel: xfs_add_to_ioend+0x124/0x219 [xfs] Jun 22 07:27:13 MikuServer kernel: xfs_do_writepage+0x346/0x483 [xfs] Jun 22 07:27:13 MikuServer kernel: ? clear_page_dirty_for_io+0x161/0x19f Jun 22 07:27:13 MikuServer kernel: write_cache_pages+0x25f/0x326 Jun 22 07:27:13 MikuServer kernel: ? xfs_add_to_ioend+0x219/0x219 [xfs] Jun 22 07:27:13 MikuServer kernel: xfs_vm_writepages+0x50/0x6f [xfs] Jun 22 07:27:13 MikuServer kernel: do_writepages+0x28/0x51 Jun 22 07:27:13 MikuServer kernel: __writeback_single_inode+0x2a/0x150 Jun 22 07:27:13 MikuServer kernel: writeback_sb_inodes+0x1c9/0x351 Jun 22 07:27:13 MikuServer kernel: __writeback_inodes_wb+0x63/0x9a Jun 22 07:27:13 MikuServer kernel: wb_writeback+0x10f/0x198 Jun 22 07:27:13 MikuServer kernel: ? wb_workfn+0x144/0x1e3 Jun 22 07:27:13 MikuServer kernel: wb_workfn+0x144/0x1e3 Jun 22 07:27:13 MikuServer kernel: process_one_work+0x155/0x237 Jun 22 07:27:13 MikuServer kernel: ? rescuer_thread+0x275/0x275 Jun 22 07:27:13 MikuServer kernel: worker_thread+0x1d5/0x2ad Jun 22 07:27:13 MikuServer kernel: kthread+0x111/0x119 Jun 22 07:27:13 MikuServer kernel: ? kthread_create_on_node+0x3a/0x3a Jun 22 07:27:13 MikuServer kernel: ? call_usermodehelper_exec_async+0x121/0x128 Jun 22 07:27:13 MikuServer kernel: ? kthread_create_on_node+0x3a/0x3a Jun 22 07:27:13 MikuServer kernel: ret_from_fork+0x35/0x40 Jun 22 07:27:13 MikuServer kernel: Code: 0f c1 15 a4 00 ed 7e 81 fa ff 0f 00 00 7e 0d 48 89 04 24 e8 5b ff ff ff 48 8b 04 245a c3 48 8b 47 48 48 8d 57 48 48 39 c2 74 02 <0f> 0b 48 89 fe 48 8b 3d f0 4e b5 00 e8 2e 06 fc ff 65 ff 0d 64 Jun 22 07:27:13 MikuServer kernel: RIP: free_buffer_head+0xd/0x3f RSP: ffffc90001a17338 Jun 22 07:27:13 MikuServer kernel: ---[ end trace d089eb21f4d110c3 ]--- I'm running into an issue where about 10 minutes after I start a transfer from another machine, it completely stops and I'm not able to connect to the SMB share anymore. The above are the logs that are printed out when it happens. root@MikuServer:/var/log# free -h total used free shared buff/cache available Mem: 15G 186M 274M 466M 15G 14G Swap: 0B 0B 0B It seems like memory is still being used by the cache? (Shouldn't this be flushed to disk) The above are transfer statistics for the first few minutes. If I try to stop the array, it gets stuck unmounting /mnt/user and I'm unable to kill the smbd thats holding files open on that path. Any information would be appreciated. Hardware Specs: Intel i3 8100 16gb DDR4 GIGABYTE H370N Dell H310 Disks: 3x 6TB WD Red Pro 512GB Samsung 970 PRO (NVME) 256GB Samsung 860 PRO (SATA) All in a SilverStone DS380B

I tried connecting it directly to the breakout cables and it still has the same issue.

Hardware specs: https://www.amazon.com/gp/product/B07CL2STVS mini-SAS -> Sata cables Intel i3 8100 16gb DDR4 GIGABYTE H370N IBM Lenovo ServeRAID M1215 (equivalent to LSI SAS9340-8i) Disks: 3x 6TB WD Red Pro 3x 3TB WD Red Pro 512GB Samsung 970 PRO (NVME) 256GB Samsung 860 PRO (SATA) All in a SilverStone DS380B The first issue I ran into was a SMBUS issue where I had to tape off pins 5 and 6 in order for my system to even start POSTing. Without it, the fans would spin on full blast and there would be no display output. Once that was solved, I ran into another issue where my motherboard would hang on POST with the typical POST screen, but only displays "Press DEL to enter setup". If DEL is hit, "Entering Setup" appears, but nothing happens. I don't have any way to figure out the post code since theres no onboard LED display, but it only occurs when theres more than one drive attached to a single SFF-8643 port. I tried many different drives (combination of physical/ssd) and was unable to get it to boot. Actually while I was writing this message, it finally finished the POST, but only one disk is visible in the BIOS. It seems like its hitting a 5 minute timeout or something similar before it skips over the drive and moves onto the next? Previously I ran this backplane with a Supermicro AOC-SAS2LP-MV8 HBA, but its performance was horrendous during a parity check, so I don't believe its the backplane, but there could be a compatibility issue with the two. Update: Enabled Legacy Bios in the UEFI menu, then forced the Legacy BIOS under the CSM support settings, and I could get the card to boot in a reasonable time. Attached is a screenshot of the syslogs that appear when the drives are being detected. The logs are with two disks attached. Thanks for any information you can give.