JPW

A few updates on my testing: As there did not seem to be a cable problem, I swapped the Sata port 1 and 2 cables back to match the wiki build document. Note: the UPS battery is not connected. I did attempt to downgrade the Seagate Barracuda SSD 120 1TB to 3.1 specifications. The downgrade failed on an error so that avenue has been abandoned. I acquired a new Samsung SSD 860 EVO 1 TB hoping that it works better than the Cuda 120 SSD. I retested 5 different drives on each Sata port. The drives were WD RED 10TB #2 (two Red 10TB drives tested are labeled #2 and #3, just the number on the drive rail) Samsung SSD 86 EVO - 1 TB SanDisk Extreme II SSD - 240GB Seagate Barracuda SSD 120 - 1TB WD RED 10TB #3 For each Sata port, the a single drive was installed with a power on boot to determine whether the drive was visible/recognized after the kernel was started. port 1: recognized: 2, 5 not recognized: 1, 3, 4 port 2: recognized: 1, 3 not recognized: 2, 4, 5 port 3: recognized: 1, 3 not recognized: 2, 4, 5 port 4: recognized: 1*, 2*, 3, 5 not recognized: 4 port 5: recognized: 1, 2, 3, 4, 5 not recognized: * - very slow boot, many resets, recognized at 3.0Gbps An observation: during the u-boot process there is a pause just after "Net: eth0: ethernet@fe300000" and before "scanning bus for devices..." of about 10-12 seconds With a drive installed in Sata ports 3, 4, 5, front panel LED lights up almost immediately after "Net: eth0: ethernet@fe300000" is displayed. If the LED is not on, the drive will not be recognized. On Sata port 1 and 2, there is a 10-12 second pause after "Net: eth0: ethernet@fe300000" is displayed and then the bus scan. I'm assuming this is a staggered power start up, Sata ports 3, 4, 5 and then Sata port 1, 2. Just visually, it seemed more random on port 1 or 2 whether a device has time to complete power up before the bus scan. I'm looking for a 3 drive configuration; a 1 TB SSD with the two WD RED 10 TB HDD. As the Cuda SSD 120 only seems to be recognized in port 5, I'll proceed with the Samsung SSD 860 EVO. But what ports? Ports 4 and 5 seem the most reliable so I'm putting the HDDs on these ports. My preference was to have the SSD on port 1 but I did experience a couple of times where the SSD wasn't recognized in this 3 drive configuration. So, I'm testing with the SSD on port 3 and the HDDs on port 4 and 5. So far, this is looking to be the best configuration. It's odd that the SSD was not recognized on port 3 during the single drive tests but seems stable with three drives. Jim

I installed Debian/Armbian 5.8 image from Kobol wiki: Armbian_20.08.21_Helios64_buster_current_5.8.17.img.xz I swapped the cables for Sata 1 and Sata 2 ports. The SSD drive was not recognized on either port. So, back to basics. Using a single WD Red 10TB HHD (factory fresh: no portioning, no formatting), I tested each slot. The WD HHD was recognized on all 5 Sata ports. So, not a cable problem. Repeating the process with the Seagate Barracuda SSD 120 drive (factory fresh), the SSD is not visible on ports 1, 2, 3, and 4 but is visible on Sata port 5. With a 2nd factory fresh Seagate Barracuda SSD 120 drive, I repeated the process. Again only visible on Sata port 5. Using a old used SanDisk SDSSDXP240G SSD, I repeated the process on port 1 to 4 and it was visible on all ports. So, it seems that there is an issue with Seagate Barracuda SSD 120 drives being only visible on Sata port 5, with the shortest cable length. But, could I just have 2 bad drives? So, I connected one of the SSDs to a USB bridge to test against my Windows desktop. I could see the drive so I partitioned and formatted the drive. Seems to be OK. So I moved the SSD back to the Helios64 (don't remember what port) and it was now visible but in a degraded 3Gbps link speed. So, I removed the Windows partitions and relabeled and partitioned the drive. Now on Sata port 1, I can see drive with the following messages being generated on boot: At least, the Cuda 120 SSD is being recognized. One Seagate troubleshooting article indicates: Since, the old Sandisk SSD worked on port 1-4, I might try downgrading the Cuda 120 SSD to the 3.1 specs (if I can find the hardware in my collection of bits). So, some progress. But, with all the drive plugging/unplugging, my Helios64 box has lots of pink plastic shavings from the drive rails. Jim

I just received my Helios64 system and built it today. The build included 3 drives, a SSD in port 1 and two HHD's on port 2 and 3. After the build, I installed Debian Kernel: 5.8 on an micro SD card. The system booted but only showed the drives in port 2 and 3. I have moved the drive in port 1 to port 4, rebooted and the SSD drive shows up. Also, a HDD in port 1 does not show up: no blue Light on the front panel and not visible in "fdisk -l" . The output from ' dmesg | grep "ata" ' results in: with the 'ata1: SATA link down (SStatus 0 SControl 300)' message repeating at least every second with a drive on Sata port 1. I dismantled the system to unplug and replug the Sata port 1 cable but no change. Any help to resolve my problem is appreciated. Thanks, Jim