Some signers have reported that they cannot write their MockSignatures to Stackerdb. When I looked into it, I found that they were trying to write to an off by one slot id (e.g. 22 when it should be 21). Somehow the signer slot id is off by one so there must be something wrong in the reconfiguration calls or in the parsing of signer slots.
Strangely this signer was configured correctly, however during reward cycle rollover...the newly activated reward cycle signer was off by one. When the signer was killed and restarted, it configured itself correctly.
EDIT: Slot id was wrong because the signer configured itself BEFORE the reward set was calculated, reading the old reward cycles' stackerdb. (i.e. cycle N+2 was loading cycle N's reward cycle info)
Some signers have reported that they cannot write their MockSignatures to Stackerdb. When I looked into it, I found that they were trying to write to an off by one slot id (e.g. 22 when it should be 21). Somehow the signer slot id is off by one so there must be something wrong in the reconfiguration calls or in the parsing of signer slots.
Strangely this signer was configured correctly, however during reward cycle rollover...the newly activated reward cycle signer was off by one. When the signer was killed and restarted, it configured itself correctly.
EDIT: Slot id was wrong because the signer configured itself BEFORE the reward set was calculated, reading the old reward cycles' stackerdb. (i.e. cycle N+2 was loading cycle N's reward cycle info)