oldguy

Good day. To verify all your information i did a test yesterday, but before that we can discard all the network issues, i have no network packet drops, the link is stable. First test Local network transferring 41 GB file full throttle 110 Mb/sec, the transfer dies in 1-2 minutes. Second test Winscp transfer over the local network directly to drive without rate limiting 80Mb/sec dies same 1-2 min Third test Winscp with rate limit to 20Mb/sec transfer never dies and finished. With this information which confirms your say about cache, however instead of slowing down, it just hangs. As i have tested the transfer via SCP and SMB with the comparable results i suppose we can assume the issue is not in underlyeing protocol but something deeper and related to software. Lets say i take one HDD out, and put it directly into USB enclosure, i think that i will able to transfer files to that Hard Drive without any issue, and it will perform based on it's specification. In this case why i am not able to transfer files via SMB/SCP to a share at maximum speed possible, and need to rate limit my transfer ? Thank you

I have 4 x TB SSD (pretty empty) (tested in raidz before), and 4 6TB HDD 2 parity 2 data. I dont really understand your information, i am the only user of unraid, and loading a test file of 41 GB, the speed rate over vpn is 40mb/s at start. When i transfer a file to cache / or to HDD's, you say they are not able to handle that transfer speed load so they drop speed and cut the transfer ? Why the system do not decrease the transfer speed ?

Good day, i am facing a problem related to SMB transfer failing consistently 1: When i transfer files via local network 2: When i transfer file wian Wireguard VPN The problem is allways the same, a good transfer rate for 4-5 min, than drop to minimum / zero. Demonstration video Attaching here diagnostics file just after the transfer test. I have tried a network disk attached or directly to share, it's the same. Also i have tried to transfer to Cache or Array directly, no difference. Tried multiple PC's, / different users / shares same result. During transfer network ping to unraid server remains stable and i dont have any network loss Any ideas what should i do ? tower-diagnostics-20240416-1708.zip

Yes i have a wifi connection to the Wifi6 AP, unraid is on the cable network. the issue is affecting pretty much everything for example my acronis backup was 15 megabits/s, after the ip manual resolving i am getting I have configured the backup solution to use the IP path, so next time i wont have this problem to solve, before that it was using smb path.

Good day so the problem is interesting, when i use this adress \\TOWER\home\admin\ i am getting 5-30 megabits per second, with fluctuating speeds Then i open the same share via IP adress, \192.168.6.50\home\admin and i am getting near full speed of iperf around 400-450 Mbit/sec Then, if i go back using a SMB names, i get the same perf as when using IP adress, but i need to do it only in the order i have described. Here is my iperf i did with the server from my PC I have tried to reboot the unraid os, upgrade it to latest without any effect. Currently i dont have information how much time the "Good speed" will remain active. Have someone encounteres such issue ? It think it is related to SMB protocol, i am using windows 11. Thank you PS. Attaching diagnostics tower-diagnostics-20230727-1500.zip

Good day, quick question, i have 4x1TB drives for cache, they are configured as mirror: I see two sets devices in mirror mode which means that i am getting something like raid 10 (2x write performance 4x read ?) Thank you

I have rebooted, well i suppose it teaches me to have another backup of backup )

This is getting stranger an stranger, 2 of my 4 array disk decide to go down I was trying to plug in an external USB hard drive during that process, have you encountered such behaviour ? Rebuilding is in process 9 hours left. All 4 disk are on the same port of the sas controller flashed in IT mode, i dont think it's a connector

Hello, i have expanded array with a one M.2 256 gb disk i have laying around, and created a share which only will use this disk, i have a stable transfer rate around 40mb/s which is to be expected with such disk, the transfer of 240 GB file finishes fine. When i transfer with old array with /or without cache enabled, i got transfer fluctuations, at first i am at 100MB then dropping to 0, then up and down. After some ~100gb time the transfer time outs. Do you think if i buy 2 more 1 tb drives to make 4 x SSD 1TB RAID 10 cache, it will be better ? I dont remember if i can do raid 10 with cache. Thank you

Good day, thank you for your reply, attaching diagnostics, last attempt was over scp, before that i have tried to transfer without cache via smb tower-diagnostics-20230413-1801.zip

Good day, my current setup is simple 4x 6TB 5900 rounds spin disks 2 parity 2 storage 2x1TB SSD caché disk in "Mirror" I was able to transfer my files of different sizes with cache enabled, the speed transfer was 100MB/S when writing to RAM, then dropping to 0 intermitenly, the larges file was 80 GB. Right now i am facing a problem, i have 3 200 GB backup files i am not able to transfer, the File Explorer fails at 50% The transfer fails when there is a long time when the transfer stops at 0mb/s I have tried to disable/enable cache and write directly to array, it fails, i have tried to use SCP instead, it fails. From this point i dont really know if there is any way i can transfer files using unraid. Is there any mode which enables minimal speed so the explorer do not fail ? Thank you