rockrimmon

Members
  • Posts

    6
  • Joined

  • Last visited

Recent Profile Visitors

The recent visitors block is disabled and is not being shown to other users.

rockrimmon's Achievements

Noob

Noob (1/14)

1

Reputation

1

Community Answers

  1. Problem resolved. The ASRock Rack X470 motherboard requires manually forcing a split between two of the PCIe slots when using two x8 PCIe devices under specific conditions that my setup required. I updated the setting in the BIOS and both HBAs are now seen and all data and cache drives are found and array started error free. I may replace the two 8 channel SAS/SATA HBAs with a single 16 channel device to make things a bit cleaner. Thanks for the assistance.
  2. Sorry for the typos. I reused the HBAs as well as the breakout cables from the HBAs to the drives. I replaced the power supply and used the new power supply cables, none of the old power supply cables were used. As noted in my prior response, it looks like the PCIe HBAs may not be working. I will investigate that further tomorrow.
  3. Attached is the diagnostic zip file. In looking deeper into the details, it looks like no media (HDD nor SSD) is being detected at all. The PCIe slots on the new motherboard are numbered different than the old SuperMIcro, not sure if that is causing any issues. No disks are showing as unassigned, nor as available, and no disks are found in the array view. I am wondering if my HBAs are working, I will try stripping out everything but one HBA and then try attaching drives to the one HBA and see if anything changes. lv426-diagnostics-20221222-2012.zip
  4. Hello, I had my Unraid server running with all disks (7 hard drives and 6 SSDs) configured and running. All drives were connected via two PCIe HBAs. The old server was a SuperMicro motherboard and Xeon processor. The motherboard failed suddenly with no errors prior to failure. Applying power resulted in no power on self test running at all,. fans started, but nothing else worked. After verifying the power supply was good I started looking for replacement parts. The SuperMicro was not available on Ebay at the time, and I had found that the Xeon CPU was running pretty much maxed out at all times. I ended up with a new ASRock Rack X470D4U2-2T motherboard with an AMD Ryzen 9 5900X 12-core, 24-Thread processor. I rebuilt the internals with the new motherboard and AMD processor, I re-used the HBAs and all cabling from the HBAs to the drives is the same. Upon booting the new setup Unraid starts up with no problems for the OS. However, every drive (parity, data, cache) is coming up as No Device on the Main tab of the Unraid UI. I know New Config can be used for some array issues. But before I used that big hammer I wanted to check if that was the appropriate approach in this scenario or if anyone had any other thoughts to try? Thanks,
  5. Hello, I have multiple Mac systems that I would like to use this Docker for. Would I be correct if I wanted each Mac to have its own TimeMachine backup area would I add multiple paths like the following: /mnt/user/timemachine/Mac1 /mnt/user/timemachine/Mac2 /mnt/user/timemachine//Mac3 Etc. Any other changes I would need to make in the Docker install? Thank.
  6. I just finished building an unRAID server and I am wondering if my intention of how to use the server is fundamentally flawed? My intention was to make the server the target for four Mac computers for Time Machine backups and/or third party backups (I have tried Carbon Copy Cloner, SuperDuper!, Chronosync, Yoyotta and simple Finder drag and drop so far in addition to Time Machine). I also copied from an external USB drive 12TB of all of my photography (connected the external drive direct to the server and used Krusader to copy from the USB drive to the server). As seen below the equipment in the server should be fairly robust. All drives are connected to two LSI PCIe adapters, the HDDs connected to one of the LSI boards, the cache on the second LSI PCIe board. The server has been rock solid, it has been up over 8 days since first power on after the build, no crashes, no reboots, zero errors on any disc. Drives are running typically in the mid 30 to low 40 degrees C, hottest ever was a drive that got to 48 degrees C (I am changing the cooling in the case to address that). CPU has never gone over 38 degrees C. unRAID is version 6.9.2. What I have discovered is that trying to copy the entire contents of any of my Macs to the server, whether via Time Machine, or any of the other alternatives mentioned above is abysmally slow, average transfer rates are ranging between 300Kb/s to 240Mb/s with most of the network transfers in the less than 2Mb/s range, these tests are all on a 10GbE network, to the point of not even being a viable option. I was prepared for the first Time Machine backup to take a long time, although 40+ hours was beyond what I expected for less than 1TB of data on the client. The issue with Time Machine is that every subsequent incremental backup is taking more than an hour in general, thus hourly backups cannot happen effectively. For any of the other options I have tried even after 48+ hours none of the other options have even reached 50% completion of a source machine with just 700 GB to be copied. In fact, Chronosync (which normally is quite speedy to a locally attached external drive, even standard old USB 2.0 is usable) I have had running over 24 hours to just copy the Applications folder of one of my Macs to the unRAID server as a test. This Application folder has 80GB in it, and after 24 hours is nowhere near complete (based on the number of Applications already copied to the server, I would say less than 50% of the 80GB Applications folder has been copied in over 24 hours). I have tried two different clients, an iMac Pro and a M1 MacBook Pro. The iMac is using its internal 10GbE connection, and I have a Thunderbolt to SFP+ 10GbE adapter on the M1 MBP, and the server is connected via SFP+ DAC from the 10GbE adapter in the server to the same 10GbE switch the iMac Pro and M1 MBP are connected to. I have applied all of the tuning suggestions in Spaceinvader One's videos, I have verified flow control is disabled. CPU utilization on the server has never exceeded 51%, and in most cases is in the 20% or less range, memory utilization is right at 11% consistently, I have all shares set to use cache, the cache pool is two Crucial MX500 1TB SSDs. I have run iPerf3 in all combinations of what machine is client versus server and consistently see 9.3Gb/s, or higher, so I know the network appears to be solid. I created a 10GB single file filled with zeros and it transfers at well over 2Gb/s from client to server. But any attempt to do a full system backup, where there are 1000's of small files crawls so badly as to be unusable. So, the question is, was I just way out of line with my expectations of being able to use the unRAID server as a target for hourly incremental backups (assuming I can even get the initial full backup to ever complete)? I am wondering if I should have gone with TrueNAS, but I really like the UI of unRAID and the apps and dockers that are available. Based on what I am seeing with respect to files hitting the cache first, and cache never reaching capacity, and given the server performance numbers, I suspect that even going to something like TrueNAS will not significantly change anything, other than the multiple spindles to write to given the distributed parity in TrueNAS and ZFS, I am just not sure if that is really where the bottle neck is, or if this is just a fundamental limitation of the reality of trying to backup numerous small files with a NAS as the target using SMB? The diagnostic files have been attached to this post. Appreciate any ideas, or verification if anyone is using an unRAID server as I am attempting to? The server equipment is as follows: Supermicro X9SCL/X9SCM, Version 1.11A Intel® Xeon® CPU E3-1270 V2 @ 3.50GHz 32 GB DDR3 ECC memory Ethernet controller: Intel Corporation 82599ES 10-Gigabit SFI/SFP+ Network Connection Parity 1 ST18000NM000J-2TV103_ZR509DE8 - 18 TB (sdb) Parity 2 ST18000NM000J-2TV103_ZR503DR8 - 18 TB (sdc) Disk 1 ST18000NM000J-2TV103_ZR503L20 - 18 TB (sdd) Disk 2 ST18000NM000J-2TV103_ZR503NPW - 18 TB (sde) Disk 3 ST18000NM000J-2TV103_ZR502BLR - 18 TB (sdf) Cache CT1000MX500SSD1_2152E5F7C26A - 1 TB (sdg) Cache 2CT1000MX500SSD1_2152E5F7C283 - 1 TB (sdl) Part of cache pool Serial Attached SCSI controller: Broadcom / LSI SAS2308 PCI-Express Fusion-MPT SAS-2 (rev 05) Serial Attached SCSI controller: Broadcom / LSI SAS2308 PCI-Express Fusion-MPT SAS-2 (rev 05) Share configuration: appdata Prefer : Cache DaveDownloads Yes : Cache DaveiPhoneVideos Yes : Cache DaveiPhotoExports Yes : Cache DaveMacStudioClone Yes : Cache DaveMacStudioTimeMachine Yes : Cache DaveMBPClone Yes : Cache DaveMBPTimeMachine Yes : Cache DaveMovies Yes : Cache DavePictures Yes : Cache DaveWebSites Yes : Cache domains Prefer : Cache isos Yes : Cache system Prefer : Cache lv426-diagnostics-20220504-0836.zip