So I’ve been mining for a couple years now, and just moved all of my miners over to HiveOS as of 2 weeks ago. Last night I received a “NAND flash error detected” notification on 2 of my L3+'s for the first time, and I’m not sure how to handle this. Error message details…
"Mining is not affected (most likely). It could be in connection with an unclean reboot OR, rarely, with a NAND wear and tear. The first is completely harmless, but the second does suggest that the ASIC’s control board may need a service or replacement in the very near future.
Errors:
[ 4.377471] jffs2: jffs2_scan_inode_node(): CRC failed on node at 0x00fc87d8: Read 0xffffffff, calculated 0x21eb2a22".
Now, I have 3 plausible ideas of what this could be, but I’d appreciate any input from other miners in the community who have experienced this for themselves.
Possible virus [(thinking unlikely) but was using unsigned version of BlissZ before move to HiveOS]
Failing control boards on the 2 L3+'s (seems pretty coincidental)
Power outtage damaged control boards [(Please god no) from sudden shut off then restart, maybe?]
I am thinking of re-flashing the firmware onto these two L3+'s via MicroSD, like I do with all of my miner’s but I don’t want the downtime if its not going to solve the issue. Additionally, I have rebooted both miners several times and still receive the message.
Any knowledge on the matter or ideas would be greatly appreciated! Thanks!
I received the same message couple times right after sudden power outage. So it really might be the " in connection with an unclean reboot" and maybe there is nothing to worry about if it only shows after the reboot.
Very interesting as I have the same history with Blissz prior to HOS. Via SSH looked at files and 4 files in the cgi-bin get overwritten by the Blissz virus which is somewhere else on the system. If you (prior to HiveOS, with Blissz still on miner) do a grep for “stratum” under root/www/pages, you will see a set of code that contains other than your pool or antminer pool (pass =123, which is a default Bitmain entry). Anything else is suspect as dev or virus injected code. But I also notice a while after reset with Hive OS that there is the same problem that my hash rate falls until reflash, which fails soft flash and has to be done via SD. So I think we have the same virus, and should let it be known that Blissz carries this on likely all copies in the wild.
After HOS is put on, I don’t find the same changes, but it must be affecting something else.
-Dan
Good to know someone else had this error after a power outtage. I continually received the error message for about 10 days if i had to reboot, than suddenly, poof it went away and i haven’t seen it since. So i would agree, it doesnt seem like much of a worry.
That is the exact problem i had with BlissZ when i was using it. The only difference in my case was one day i went to access my miners through IP and my password had been changed. Thats when i found out some versions of blissz have a virus that will move through your network and infect other miners. Long story short, i run HiveOS because of how terrible blissz was. But i avoid blissz or any unsigned firmware at that. Lost a days worth of mining on all of my rigs figuring that lesson out the hard way.
You reminded me same happened with Blissz also. SSH was not working at first on Blissz, but finally I got it to work one day for no apparent reason. I had to use the NAND jumper to flash also as the virus prevented SD card flashing. Like you said, a lot of time wasted. Here are the cgi-bin lines and the CGI-BIN files where they were located. No idea who this account or pool is other than it is likely the pirate or dev trail.
