Pages:
Author

Topic: Innosilicon releases A9 Zmaster 50ksol/s Equihash miner at 620W - page 39. (Read 45686 times)

newbie
Activity: 45
Merit: 0
My one hashcard is not working. Other 2 working good. Now I will send back one card to them and they should fix it.
newbie
Activity: 46
Merit: 0
My one GPU module on A9 stopped working. Any troubleshoot ideas? Support is not responding.
Now mining on 2 GPU modules at average 30 Ks on nicehash.
Sad to loose 1/3 of capacity.

It's the same problem with my second A9. I am mining at 25 - 30 ks with that one, it cannot be changed to another A9 because the local seller told me that A9 now is not available in his stock. Sad to you and me.
newbie
Activity: 25
Merit: 0
they provided no explanation...pretty much just said no, sorry, wont give it to you, neener neener neener.

ive started looking at brute forcing the ssh password knowing that some keywords will probably be "inno" "t1t2t3" "a9" based on previous innosilicon passwords but probably wont have much luck....

Is it possible to mount the firmware on a linux box or something, so could try and brute force it a little more firmly Smiley

Yes, it's possible!  Here is the /etc/shadow entry from the firmware file if anyone wants to take a go at it:

root:$6$AFqSg1Lk$PcjGWzIPX1ghz7GqGzXoFq0WsYTAwz5VsbG/DiKBQmxYno4igKlPLhqwTkdenTevxS9K9bFqzUMGiB1KdIX3G.:10933:0:99999:7:::

Please share any results here!


I'm trying to figure out the ssh password to my Inno D9 and it looks like you guys are going down the same paths as I have (without success). I don't think it's reasonable to brute force this. I tried it with hashcat for about a week and didn't find anything. My latest theory is to modify the SWU file as it contains the rootfs.ubi file. Change the hashed password in the binary that's in the shadow file to a known hash value (with the particular salt), then update the hash of the rootfs.ubi section. Unfortunately, I'm stuck at being able to reproduce the hash of the original .ubi file section. Then, flash the modified SWU file. However, I'm really trying not to brick my miner. I just haven't dealt with SWU files before so I'm a little cautious here.

I'm interested in learning a bit more about this. How did you get the /etc/passwd? I've extracted the .swu but just see BOOT.bin and rootfs.ubi etc.  Do we know which hashing algorithm that password is? MD5 or SHA512

Which part of the root:$6$AFqSg1Lk$PcjGWzIPX1ghz7GqGzXoFq0WsYTAwz5VsbG/DiKBQmxYno4igKlPLhqwTkdenTevxS9K9bFqzUMGiB1KdIX3G.:10933:0:99999:7:::

are we trying to force?  I guess if there's a salt as well there's little chance of figuring it out.

The SWU file is several files in one, each with a section and hash of the data of that file. If you search for "$6$", you'll find that contents of the shadow file which has that root:$6$... structure. The $6$ indicates it's a SHA512. The salt is the characters in between the second and third "$" signs. The hashed portion of the password is after the third "$" sign. Note that these are put into a different format due to linux's "crypt" routine. Read here:  http://manpages.ubuntu.com/manpages/trusty/man3/crypt.3.html

This is why brute-force cracking this is not reasonable (in my mind). I think it's more probable that the SWU image gets updated to a known hashed password.

I'm going to throw 8 1080Ti at it for a while trying to brute force crack it.
newbie
Activity: 141
Merit: 0
Looks like there's a new firmware on their download page. 0718  Perhaps it will have performance improvements Smiley

http://www.innosilicon.com/html/support_en/download.html

I tried it earlier today for about 6 hours.  No obvious improvements and in fact, the hash rate was very slow to ramp up - taking over an hour to plateau.  I went back to the 7/6 firmware.


Running a test for new firmware on Nicehash and Flypool. At this moment looks better than 6/30 and 7/6... will provide updates in 12 hours.
newbie
Activity: 25
Merit: 0
Looks like there's a new firmware on their download page. 0718  Perhaps it will have performance improvements Smiley

http://www.innosilicon.com/html/support_en/download.html

I tried it earlier today for about 6 hours.  No obvious improvements and in fact, the hash rate was very slow to ramp up - taking over an hour to plateau.  I went back to the 7/6 firmware.
jr. member
Activity: 36
Merit: 8
Looks like there's a new firmware on their download page. 0718  Perhaps it will have performance improvements Smiley

http://www.innosilicon.com/html/support_en/download.html
jr. member
Activity: 55
Merit: 5
they provided no explanation...pretty much just said no, sorry, wont give it to you, neener neener neener.

ive started looking at brute forcing the ssh password knowing that some keywords will probably be "inno" "t1t2t3" "a9" based on previous innosilicon passwords but probably wont have much luck....

Is it possible to mount the firmware on a linux box or something, so could try and brute force it a little more firmly Smiley

Yes, it's possible!  Here is the /etc/shadow entry from the firmware file if anyone wants to take a go at it:

root:$6$AFqSg1Lk$PcjGWzIPX1ghz7GqGzXoFq0WsYTAwz5VsbG/DiKBQmxYno4igKlPLhqwTkdenTevxS9K9bFqzUMGiB1KdIX3G.:10933:0:99999:7:::

Please share any results here!


I'm trying to figure out the ssh password to my Inno D9 and it looks like you guys are going down the same paths as I have (without success). I don't think it's reasonable to brute force this. I tried it with hashcat for about a week and didn't find anything. My latest theory is to modify the SWU file as it contains the rootfs.ubi file. Change the hashed password in the binary that's in the shadow file to a known hash value (with the particular salt), then update the hash of the rootfs.ubi section. Unfortunately, I'm stuck at being able to reproduce the hash of the original .ubi file section. Then, flash the modified SWU file. However, I'm really trying not to brick my miner. I just haven't dealt with SWU files before so I'm a little cautious here.

I'm interested in learning a bit more about this. How did you get the /etc/passwd? I've extracted the .swu but just see BOOT.bin and rootfs.ubi etc.  Do we know which hashing algorithm that password is? MD5 or SHA512

Which part of the root:$6$AFqSg1Lk$PcjGWzIPX1ghz7GqGzXoFq0WsYTAwz5VsbG/DiKBQmxYno4igKlPLhqwTkdenTevxS9K9bFqzUMGiB1KdIX3G.:10933:0:99999:7:::

are we trying to force?  I guess if there's a salt as well there's little chance of figuring it out.

The SWU file is several files in one, each with a section and hash of the data of that file. If you search for "$6$", you'll find that contents of the shadow file which has that root:$6$... structure. The $6$ indicates it's a SHA512. The salt is the characters in between the second and third "$" signs. The hashed portion of the password is after the third "$" sign. Note that these are put into a different format due to linux's "crypt" routine. Read here:  http://manpages.ubuntu.com/manpages/trusty/man3/crypt.3.html

This is why brute-force cracking this is not reasonable (in my mind). I think it's more probable that the SWU image gets updated to a known hashed password.
jr. member
Activity: 36
Merit: 8
they provided no explanation...pretty much just said no, sorry, wont give it to you, neener neener neener.

ive started looking at brute forcing the ssh password knowing that some keywords will probably be "inno" "t1t2t3" "a9" based on previous innosilicon passwords but probably wont have much luck....

Is it possible to mount the firmware on a linux box or something, so could try and brute force it a little more firmly Smiley

Yes, it's possible!  Here is the /etc/shadow entry from the firmware file if anyone wants to take a go at it:

root:$6$AFqSg1Lk$PcjGWzIPX1ghz7GqGzXoFq0WsYTAwz5VsbG/DiKBQmxYno4igKlPLhqwTkdenTevxS9K9bFqzUMGiB1KdIX3G.:10933:0:99999:7:::

Please share any results here!


I'm trying to figure out the ssh password to my Inno D9 and it looks like you guys are going down the same paths as I have (without success). I don't think it's reasonable to brute force this. I tried it with hashcat for about a week and didn't find anything. My latest theory is to modify the SWU file as it contains the rootfs.ubi file. Change the hashed password in the binary that's in the shadow file to a known hash value (with the particular salt), then update the hash of the rootfs.ubi section. Unfortunately, I'm stuck at being able to reproduce the hash of the original .ubi file section. Then, flash the modified SWU file. However, I'm really trying not to brick my miner. I just haven't dealt with SWU files before so I'm a little cautious here.

I'm interested in learning a bit more about this. How did you get the /etc/passwd? I've extracted the .swu but just see BOOT.bin and rootfs.ubi etc.  Do we know which hashing algorithm that password is? MD5 or SHA512

Which part of the root:$6$AFqSg1Lk$PcjGWzIPX1ghz7GqGzXoFq0WsYTAwz5VsbG/DiKBQmxYno4igKlPLhqwTkdenTevxS9K9bFqzUMGiB1KdIX3G.:10933:0:99999:7:::

are we trying to force?  I guess if there's a salt as well there's little chance of figuring it out.
jr. member
Activity: 55
Merit: 5
they provided no explanation...pretty much just said no, sorry, wont give it to you, neener neener neener.

ive started looking at brute forcing the ssh password knowing that some keywords will probably be "inno" "t1t2t3" "a9" based on previous innosilicon passwords but probably wont have much luck....

Is it possible to mount the firmware on a linux box or something, so could try and brute force it a little more firmly Smiley

Yes, it's possible!  Here is the /etc/shadow entry from the firmware file if anyone wants to take a go at it:

root:$6$AFqSg1Lk$PcjGWzIPX1ghz7GqGzXoFq0WsYTAwz5VsbG/DiKBQmxYno4igKlPLhqwTkdenTevxS9K9bFqzUMGiB1KdIX3G.:10933:0:99999:7:::

Please share any results here!


I'm trying to figure out the ssh password to my Inno D9 and it looks like you guys are going down the same paths as I have (without success). I don't think it's reasonable to brute force this. I tried it with hashcat for about a week and didn't find anything. My latest theory is to modify the SWU file as it contains the rootfs.ubi file. Change the hashed password in the binary that's in the shadow file to a known hash value (with the particular salt), then update the hash of the rootfs.ubi section. Unfortunately, I'm stuck at being able to reproduce the hash of the original .ubi file section. Then, flash the modified SWU file. However, I'm really trying not to brick my miner. I just haven't dealt with SWU files before so I'm a little cautious here.
newbie
Activity: 25
Merit: 0
they provided no explanation...pretty much just said no, sorry, wont give it to you, neener neener neener.

ive started looking at brute forcing the ssh password knowing that some keywords will probably be "inno" "t1t2t3" "a9" based on previous innosilicon passwords but probably wont have much luck....

Is it possible to mount the firmware on a linux box or something, so could try and brute force it a little more firmly Smiley

Yes, it's possible!  Here is the /etc/shadow entry from the firmware file if anyone wants to take a go at it:

root:$6$AFqSg1Lk$PcjGWzIPX1ghz7GqGzXoFq0WsYTAwz5VsbG/DiKBQmxYno4igKlPLhqwTkdenTevxS9K9bFqzUMGiB1KdIX3G.:10933:0:99999:7:::

Please share any results here!
newbie
Activity: 62
Merit: 0
the innosilicon manual states 1000W+ PSU is needed for the A9. 

i dont understand why, because it only draws 630W.  Most PSUs are like 88-95% efficient, so a 850*88%=748W > 630W. 

performance mode doesnt do anything sooo i dont see why you really need a 1000W psu.

i bought an 850W psu for 20 bucks with a gift card to replace the high pitched whine on my bitmain psu.  i dont see why it wont work until and unless innosilicon releases new firmware to make the a9 actually perform better in performance mode.

i dont want to spend another 200$ on this miner, every $ counts for capital recovery and roi, but that high pitched whine is pretty annoying and worth 20$ to fix in my book.
newbie
Activity: 51
Merit: 0
For anyone using a standard ATX power supply please respond if my plan is correct/incorrect.

I plan on changing from the noisy power supply to a:
EVGA SuperNOVA 1000 P2 80+ PLATINUM
https://www.amazon.com/gp/product/B00EKJQM5E/ref=ox_sc_act_title_3?smid=A2Q1LRYTXHYQ2K&psc=1

If I am correct the EVGA is a single rail so it shouldn't matter which of the six vga outs I use to connect to the 3 hashboards?
I plan to use all six vga outs to connect to the hash boards.

I then plan to use a molex to pcie six pin to connect to the controller board:
https://www.amazon.com/gp/product/B01DV1Z22K/ref=ox_sc_act_title_1?smid=A1AMUYYA3CT6HJ&psc=1

I'm also getting a switch to turn the EVGA on/off. I know I could just jump it but... This is nicer...
https://www.amazon.com/gp/product/B01MSY4966/ref=ox_sc_act_title_2?smid=AQ8OIL3FGVKZE&psc=1

Will this work correctly or is there a better way? I don't want to fry the miner but this PSU is driving me to distraction!

Thanks


You might not need the molex adapter since the PSU shoud have enough connectors. Check at the EVGA page.

Thanks, your right, it has 8 six pin vga connections.

jr. member
Activity: 306
Merit: 7
For anyone using a standard ATX power supply please respond if my plan is correct/incorrect.

I plan on changing from the noisy power supply to a:
EVGA SuperNOVA 1000 P2 80+ PLATINUM
https://www.amazon.com/gp/product/B00EKJQM5E/ref=ox_sc_act_title_3?smid=A2Q1LRYTXHYQ2K&psc=1

If I am correct the EVGA is a single rail so it shouldn't matter which of the six vga outs I use to connect to the 3 hashboards?
I plan to use all six vga outs to connect to the hash boards.

I then plan to use a molex to pcie six pin to connect to the controller board:
https://www.amazon.com/gp/product/B01DV1Z22K/ref=ox_sc_act_title_1?smid=A1AMUYYA3CT6HJ&psc=1

I'm also getting a switch to turn the EVGA on/off. I know I could just jump it but... This is nicer...
https://www.amazon.com/gp/product/B01MSY4966/ref=ox_sc_act_title_2?smid=AQ8OIL3FGVKZE&psc=1

Will this work correctly or is there a better way? I don't want to fry the miner but this PSU is driving me to distraction!

Thanks


You might not need the molex adapter since the PSU shoud have enough connectors. Check at the EVGA page.
newbie
Activity: 51
Merit: 0
For anyone using a standard ATX power supply please respond if my plan is correct/incorrect.

I plan on changing from the noisy power supply to a:
EVGA SuperNOVA 1000 P2 80+ PLATINUM
https://www.amazon.com/gp/product/B00EKJQM5E/ref=ox_sc_act_title_3?smid=A2Q1LRYTXHYQ2K&psc=1

If I am correct the EVGA is a single rail so it shouldn't matter which of the six vga outs I use to connect to the 3 hashboards?
I plan to use all six vga outs to connect to the hash boards.

I then plan to use a molex to pcie six pin to connect to the controller board:
https://www.amazon.com/gp/product/B01DV1Z22K/ref=ox_sc_act_title_1?smid=A1AMUYYA3CT6HJ&psc=1

I'm also getting a switch to turn the EVGA on/off. I know I could just jump it but... This is nicer...
https://www.amazon.com/gp/product/B01MSY4966/ref=ox_sc_act_title_2?smid=AQ8OIL3FGVKZE&psc=1

Will this work correctly or is there a better way? I don't want to fry the miner but this PSU is driving me to distraction!

Thanks
newbie
Activity: 20
Merit: 0
The miner is lying at customs after reaching destination ( by DHL tracking ) for more than 11 days now. Have mailed the customs appellate, who has directed a reply but still nothing has moved. The miner was bought by crypto. Even after producing invoice of $300 (for shipping an handling ) the guy now wants bank records to be shown and not ready to accept crypto payment have started happening and there is no bank for that.  (This type of import have been discussed on this forum too and US/EU citizens have imported without much hassles).As an option I told him to return the miner to seller (which is what happens when they are not delivered) by charging whatever fee , which is also not being done. It is hoarded at the airport for 11 days now - calculate total loss from July 3rd when it got shipped !

This is what happens when a village idiot runs a country.
India brought ICEGATE ( hurriedly copied from ICE of US with shabby website ) and taxed everything flat at 30% GST (I dont know if IGST is also charged in addition) without doing an iota of work or insurance or whatever ( which probably other countries do ) . From cellphones to sims to TVs to cars to bikes to aircrafts to trains every high end electronic is imported to India - the main motive being to keep engineering / intellegentsia away from government and making a loot out of commissions from import deals. Since its colonial commonwealth genes is still rooted strong anything coming not from US/UK is going to be met with severe suspicion, hatred and food for bribe. Now since my miner was from HK and I not being a diplomat  / connected to govt in anyway they will make it more tougher.And a few people like me who don't join US outsourcing brigade which itself is a heavily brahminical (yes its a indian jewish(may not be all) problem) H1B VISA arbitrage out of cheap labor, its a nightmare.

Indian govt looks like a kid looking for the shiniest and bulky equipment , crying for it's share. This is microcosm of trade barriers, greedy and idiotic governments can harass citizens with when they feel their existence is at stake.

Don't think many are running this miner in India. The seller sent one example routed through same customs but don't know how he did receive it after 2 weeks!
jr. member
Activity: 36
Merit: 8
they provided no explanation...pretty much just said no, sorry, wont give it to you, neener neener neener.

ive started looking at brute forcing the ssh password knowing that some keywords will probably be "inno" "t1t2t3" "a9" based on previous innosilicon passwords but probably wont have much luck....

Is it possible to mount the firmware on a linux box or something, so could try and brute force it a little more firmly Smiley
jr. member
Activity: 306
Merit: 7
Maybe you could also try halong mining passwords.
newbie
Activity: 141
Merit: 0
If someone can figured out on how to get storage out from a9, it is simple to edit "passwd" file to delete current password for root from current to "nothing" and start miner again with blank password for root.

The problem is, storage needs to be connected to external terminal. I don't know which storage - flash/sdcard/etc Innosilicon using as a boot device on A9. Flash - could be an issue as it will be soldered to control board.
jr. member
Activity: 306
Merit: 7
they provided no explanation...pretty much just said no, sorry, wont give it to you, neener neener neener.

ive started looking at brute forcing the ssh password knowing that some keywords will probably be "inno" "t1t2t3" "a9" based on previous innosilicon passwords but probably wont have much luck....

Maybe someone can have a look at the firmware?
newbie
Activity: 62
Merit: 0
they provided no explanation...pretty much just said no, sorry, wont give it to you, neener neener neener.

ive started looking at brute forcing the ssh password knowing that some keywords will probably be "inno" "t1t2t3" "a9" based on previous innosilicon passwords but probably wont have much luck....
Pages:
Jump to: