id column overflow last night

Last night, at 2:12 EST, 2^31 rows had been inserted in the daily balances table, causing the next row to fail due to overflow. For several hours afterward, the mining server queued up shares in memory and allowed mining to continue, until it eventually ran out of memory and crashed.

We apologize for the problem. Mining should be back to normal. I changed the type of the column to BIGINT so that it won't overflow, and Chris will compute and adjust balances for the day so that everyone receives credit for the time they mined.

Feel free to ask any questions you might have as we continue to work on the unrelated performance improvements, which I intend to release within the week.

The server was offline for 32878 seconds, from 02:12:20 to 11:20:19. I will be correcting everyone's balances for the missing shares by increasing everyone's February 14 earnings by 61.43%, regardless of whether you were mining around that time or not. The balances have been corrected as of five minutes ago and will be paid today. Please let me know if you have any questions.