yes both the rootfs and alerts on the ram are in my log (like yours)  - though im stable and fine on btc.com at 450 Hz

My (2) DR3 are hashing 2+ days since PDU reset, but i have seen my first one that i had since monday reboot on its own a few times.  I'm holding about 8.5 - 8.7 TH/s confirme by pool side too.  About 3.5 % HW errors though.

I've got frequency maxed on both at 450, Chip temps are ranging from  I:56 O:78 on worst chip, and I54 O:72 on the best.   PCB is i:45 O:63

I see the same kernel log items you do doesnt seem to be an issue  No reboots in 2+ days though.   I'm using BTC.com pool to china, from USA (not the best idea but it was a quick one to get started with)

i clocked both of mine to the max at 450 Frequency, lots of HW errors - im at 3.6% HW error rate doing this at max frequency

From the web interface
1d6h34m24s   Current GH/S  8,526.48    AVG GH/S  8,531.99  

Both run about the same, (the other is showing 8.7 TH/s

Got 1 on monday and other on thursday (due to held up in customs).  No fees via Fedex for US customs on both (both were in same order too, dont know why 1 came monday and the other was delayed until thursday) - no power supplies ordered, but i can tell you bitmain is gaming the paperwork with random jibberish as to whats inside so fedex probably too confused on that aspect.  Different from prior antminer orders.

From the dashboard at btc.com (im in the US, using their china node so not optimized but didn't have time yet to setup anything else) as  i'm not on any pools that supported DCR stuff.  And as usual it wouldnt work at nicehash when i tried there first.

DCR earnings Yesterday
0.56099935

And its showing 17.56 TH/s avg at the pool for what it's worth and 16.5 TH/s 'real time' showing at BTc.com's pool. .  So obviously my hw errors and reported speed net down to the real number they show.

I've got too many items plugged in to tell what its pulling at the wall with the over clock sorry.  But my APC PDU shows 3.92 KW with (2) of these Antminers, and 4 x Z9 mini's and about another 200 watts for a couple of cisco ASA / switches in thatrack powered on the same pdu.  

No doubt it will be a long ROI battle at the end, but getting in as early as possible leaves some hope.

But figured that info may help some.  My opinion, its a horrible power hungry box compared to other DECRED miners coming soon, but i've learned my lesson - get the very first units, or don't bother.  Then sell it on ebay about 30-45 days in... guaranteed roi that way.   Of course just because i learned my lesson doesn't mean I will follow it, i just keep running them in the farm long past when i should sell them off usually.

One item on the cost analysis - at least if you are in the U.S., is the 25% and 2.6% tarrifs.  The Z9 will trigger the > $2500 incoming to the US tarrif check, guaranteed, and cost you a lot more.

If you buy (3) Z9 mini's at a time, per carrier (fedex/dhl/etc) and spread it out to no more than 3 at a time, you stay under the US customs $2500 import duty 'trigger'.

I know this from personal experience, with at least 6 batches of 3 x Z9 mini's over the last 35 days.  Not one has hit the 25% tarrif, 1 of the 6 hit the 2.6 % tarif (A dhl package from china vs hong kong).   I can't say with guarantee that the carriers or government wont mail me a bill 2 years from now for the 25%, but everything I've personally experienced in prior orders and the recent is stay under $2500 per 'shipment' (dont order PSU's either so its not a complete product) and you can avoid it.

Z9 full is not worth vs the z9 mini in my opinion, its cheaper for the mini's when factoring shipping tarrifs in the US at least.  But I prefer the divide and conquer, in case of hash board or other failure vs a bigger box that could all die at once.

P.s. rethink your qty (4) z9 mini plan of order, get only 3, and stay under the $2500 limit.  If you want more, order a second batch via a 2nd carrier (fedex is best, then dhl, and ups last on my list personally for using).  That will avoid things for you (based on my personal experience, i've read elsewhere others have not been so lucky as me).  but I guarantee if you exceed $2500 on any US import, you will trigger the full review and get hit.

While not alt coin 100% specific, or mining 100% specific, - i'm working on a white paper related to both so i hope the question in this spot is appropriate.

What benchmark tool (windows or linux) can anyone suggest that would provide a numeric value of output that shows both CPU and GPU that could relate to each other in the same score type.  

what I am looking for is very similar to how NiceHash algorithm checks for best output on a specific miner's computer (showing the Hash per second on the CPU vs each gpu).  The difference for me is I'm simply looking to get that type of calculation comparison in a way that may not be so cryptocurrency focused (i.e. using Nicehash as my bench mark tool).
 

This is for a white paper i'm working on that is not crypto specific but still relates.  I'm trying to find something that shows the relationship between GPU models (RX580, Vega 64, and GTX 1080 TI in my case) and any cpu to show how more efficient on power consumption the GPU is in doing a specific benchmark.  (I'm monitoring power to compare versus the 'score' of whatever benchmark I can use).

The only caveat is the benchmark tool I use needs to be publicly accepted - not home grown.  I've gone through LinPack, PassMark, BaseMark, UserBenchMark, Unigine_Heaven, and a few others.   The common thing I'm running into is they state the 'gpu' success in 1 metric (say Frames per second) versus the computer/cpu score in another metric (say flops per second).  Linpack has some gpu related items - but only for CUDA (wont work on the AMD also).  Writing my own benchmark (say a simple hash per second test say with cryptonight or something - unfortunately isn't going to meet the needs of a publicly accepted benchmark suite to use as my baseline).

My goal is to get the exact same benchmark number (say fps) compared against both to show the relationship.  I also realize that different benchmarks that are cpu specific vs gpu specific throw the results in favor of whatever they are written at.  I'm trying to visually show that a GPU is about 7 x more power efficient at the 'same' benchmarked mathematical task as say the CPU.

Sorry for the techo-babble-  but I seen some very smart posts on here and this I figured was my best shot to find one of you to point me in the write direction for my white paper research element I'm trying to produce.

thank you.

One other option.   Get some big electric extension cords / net metering and sell that 20 kw back to the grid

20kw at 12 cents per kWh is about 21,000$ a year. You won’t get that though depending on local laws you might on net metering plans

25 x l3+ antminers in free electricity of 20 kw earns about $15k right now less cost of 25 antminers ($3k used on eBay) So net $12k year one

You are better off selling the electricity to neighbors if you can back feed it some how like solar net metered services include

If you go crazy and put out in some bigger miners that cost way more up front and longer roi maybe you can out pace that for profit of just selling the the electricity

For example 60 x z9mini is about 20 kw.  That’s about $50k up front shipped.  And daily profit dropping fast I’d expect an average around 3$ a day over 12 months best case. That puts at about $1k income per miner   I know today th looks great at 10$ to 15$ daily. It’s dropping fast and avg 12 Month income
daily could be under $1 due to difficulty in a few months.  It’s a gamble you roi at all even with free electricity based on past antminer releases the last 12 months - and you have to spend a lot to find out versus that used eBay l3 antminers  I suggested is guaranteed slow grind profit. Even if litecoin dificulty grows you will roi that hardware when bought used

Cgminer.conf will auto change time stamp any time you so much as change the pool information. It is over written each time.  So they can’t see you over clocked from that file if you simply set the frequency back to 500

The log files though show your frequency each reboot and startup so delete your logs or do a factory reset from browser first and then verify logs gone on file systems

1 Word.  ok its not even a word..  HODL.

I feel the pain on the gpu (and asic) farming, we need to make mining great again.  

A year ago people said dont buy GPU/ASIC, take the money and invest it in the coin instead.  I didn't listen, I doubled down....

Well if you had followed that advice to take the same money instead of buying mining equipment, just buy coins at the prices a year ago --- you would of lost 60-70% of that investment.   At least putting it into mining hardware, you have some coin, some experience (that will help in some way some day in a career, or side project), and you can keep running it in the 'red' and HODL.  Don't bother trying to sell the gear on ebay - its like trying to sell your coin at the bottom, its just too late.  Ride it out.

if cost of the mining equipment is a major concern, go with your comments below about Asic Antminer L3+ - about $130 each on Ebay.  Don't buy them new, just buy in mass used.  They are not hugely profitable unless power cost is low (and if free, its a big win and ROI would be super fast at those costs per antminer).  For 20 KW you could place about 25 of these, each generates about $600/year right now of litecoin (but if market goes up....).   Thats a 350% plus ROI on the cost of equipment in 1 year, or break even in 3 months.  Doesn't get much better.

More expensive miners (like the Z9 and Z9 minis) look great on paper, but due to people jumping in on it (and its really new) it will have a negative ROI just like the rest of the antminers at some point within the next 6 to 12 months.  And the cost to get those new in demand ASIC (or GPU) would be top dollar, that is hard to earn back .  Buy used on EBAY (or your equivalent) and you can make easy money guaranteed with ROI versus everyone else who has to buy the latest and newest high priced gear to grind out a fast return.  With your free power you are guaranteed a win.

Good luck - tell us how it goes.

hard to believe any block chain related cloud file storage can find success, but best of luck to their team.  Having dove into Siacoin and their file storage concept for blockchain, burned enough to see that its just not the right use of the technology.

As linked just above  - quick way is to edit the using f12 via browser under your advanced settings to set the 500 value for frequency to your over clock (see https://bitcointalk.org/index.php?topic=3509006.msg45020727#msg4502072)  However another quick way depending on how your farm may be setup is just to ssh in and edit the file in /config/cgminer.conf and edit the frequency by hand there in that file, and then reboot the antminer (type 'reboot' if you are in ssh).

For those not familiar with SSH, check out a program calledWinSCP or  PuttY.  WinSCP is even easier, it lets you navigate and edit files too), and just connect using the default user 'root' default password is 'admin' to your port for SSH/SCP (port 22).  You can then if using PuttY edit the file using VI (good luck if you are new to unix, sorry but google is your friend).  WinSCP is much easier to remotely access and edit files on the antminer.  Perhaps easier than using developer mode (f12) in your browser to hand edit frequency.

Again the file is located in folder    /config    and is named cgminer.conf  - you may find it easier to manage things there on all your different antminers anyway than browser access if you have lots to manage.

p.s. I use 675 Frequency and get about 14 K up to as high as 15.92 K output - varies between my different Z9 minis for the same setting (all batch 2's) and even unstable on others.  Still adjusting to fine the best stable one.

edit: 3 of my units from 'china' (fedex delivery) clock at 675 fine and stable, produce 14 to 15.9 K output.  My other 3 units that came from Hong Kong shipping origin (DHL) all dropped out a chip after about 10-15 minutes, at 675 Clock Frequency.  at 650 Clock its stable for those 3 units producing about 13.9 K up to about 14.4 K.  Arrived within a day of each other this week batch #2 - the theory that different batches, and chip quality can differ would hold true.   Why hong kong vs china origin ship units differ I dont know.  Serial numbers on the 3 from honk kong end in 914, 919, and 932.  The China serial numbers are wildly different for letters and numbers, again all 3 arrived and shipped together from china.  Temps on all are 68C to 70C for what its worth on the over clock.  (Its hot out here and using open-air system with fans).