Miner frequently disconnects from pool

[Demontager]

I have setup public pool to run in testnet, my detailed steps were posted here https://github.com/benjamin-wilson/public-pool/issues/71

.env config

BITCOIN_RPC_URL=http://192.168.1.181

BITCOIN_RPC_USER=dem
BITCOIN_RPC_PASSWORD=mypassword
BITCOIN_RPC_PORT=5000
BITCOIN_RPC_TIMEOUT=10000

BITCOIN_RPC_COOKIEFILE=

API_PORT=3334
STRATUM_PORT=3333
NETWORK=testnet

API_SECURE=false
POOL_IDENTIFIER="public-pool"

Then i set Antminer S9 to mine test btc. The pool config at miner config page pool: stratum+tcp://192.168.1.181:3333 Worker: mrYpN6imcBNkmsceuw5C7JjtFa9AyasEBL password: x

Pool log started to show blocks, but for some reason miner getting frequent disconnects and it reconnects again by itself.

[Nest] 6983  - 11/02/2024, 11:23:17 PM     LOG [NestApplication] Nest application successfully started +20ms
API listening on http://0.0.0.0:3334
Bitcoin RPC connected
block height change
Stratum server is listening on port 3333
Killing dead clients
block height change
Killing dead clients
Killing dead clients
New client ID: : d1d2dd6f, ::ffff:192.168.1.170:42397
getblocktemplate tx count: 3548
getblocktemplate tx count: 3549
getblocktemplate tx count: 3550
getblocktemplate tx count: 3551
Killing dead clients
getblocktemplate tx count: 3552
getblocktemplate tx count: 3554
Client d1d2dd6f disconnected, hadError?:false
New client ID: : 52caab20, ::ffff:192.168.1.170:44496
getblocktemplate tx count: 3519
getblocktemplate tx count: 3522
Deleting statistics
Killing dead clients
Deleted 0 old statistics
Deleted 0 old clients
getblocktemplate tx count: 3523
getblocktemplate tx count: 3526
getblocktemplate tx count: 3527
getblocktemplate tx count: 3529
Client 52caab20 disconnected, hadError?:false
Killing dead clients
New client ID: : ba84a2f4, ::ffff:192.168.1.170:44749
getblocktemplate tx count: 3533
block height change
getblocktemplate tx count: 3520
new block
getblocktemplate tx count: 3519
getblocktemplate tx count: 3526
getblocktemplate tx count: 3529
getblocktemplate tx count: 3530
Killing dead clients
getblocktemplate tx count: 3531
Client ba84a2f4 disconnected, hadError?:false
New client ID: : 43332838, ::ffff:192.168.1.170:44785
getblocktemplate tx count: 3532
getblocktemplate tx count: 3532
block height change
getblocktemplate tx count: 3522
new block
block height change
getblocktemplate tx count: 3517
new block
block height change
getblocktemplate tx count: 3518
new block
getblocktemplate tx count: 3519

[Demontager]

Found a problem with restarts, it was related to miner settings. Watchdog log showed that miner was restarting because of low hashrate

Sun Nov  3 08:27:36 2024 Total hashrate 9609.871 is less than 90 percents of 13115.70, stopping miner
Sun Nov  3 08:34:18 2024 Total hashrate 9653.626 is less than 90 percents of 13115.70, stopping miner
Sun Nov  3 08:41:00 2024 Total hashrate 9631.782 is less than 90 percents of 13115.70, stopping miner
Sun Nov  3 08:47:42 2024 Total hashrate 9607.137 is less than 90 percents of 13115.70, stopping miner
Sun Nov  3 08:54:24 2024 Total hashrate 9470.335 is less than 90 percents of 13115.70, stopping miner
Sun Nov  3 09:08:08 2024 Total hashrate 9635.339 is less than 90 percents of 13115.70, stopping miner
Sun Nov  3 09:14:29 2024 Total hashrate 9790.512 is less than 90 percents of 13115.70, stopping miner
Sun Nov  3 09:21:11 2024 Total hashrate 9694.529 is less than 90 percents of 13115.70, stopping miner
Sun Nov  3 09:27:53 2024 Total hashrate 9782.224 is less than 90 percents of 13115.70, stopping miner

Then i've disabled restart function function

Restart if hashrate us below given number of percents of ideal hashrate

Now miner running fine without restarts, keeping 9 Th/s. But why it can't reach 13 th/s ?

[Demontager]

After investigated miner chips I found that many already yellow and red, therefore asic runs unstable and reducing hashrate. You may see that 7 Chain on my screenshot having a lot of HW errors. This ASIC is old, i bought it as second hand just for testing purposes, so it's fine.