A method begins by a first computing device generating an initial routing plan that identifies network paths for transmitting encoded data slices of an encoded data segment from the first computing device to a second computing device. The method continues with the first computing device sending a plurality of subsets of encoded data slices to network paths. Within a network path, the method continues by a relay unit determining whether the network path defined by the initial routing plan requires adjusting. When the network path requires adjusting, the method continues with the relay unit establishing an adjusted network path by at least one of adding a relay unit and deleting a relay unit. The method continues with the relay unit sending, via the adjusted network path, the corresponding subset of encoded data slices to the second computing device.

A device is provided with: memory cell array including plurality of first and second memory cells and one or more third memory cells; judging circuit that judges plurality of data values held by selected first and second memory cells of the first and second memory cells, by referring to reference potential corresponding to reference data held by a selected third memory cell; and error detection and correction circuit that detects whether or not there is error in the judged data values of the first and/or second memory cells, with judged data value of the first and second memory cells as error correcting code. When the error detection and correction circuit detects that there is error exceeding error correction capability in the judged data values, control is performed to write reference data to the selected third memory cell.

A technique is provided for accumulating failures. A failure of a first row is detected in a group of array macros, the first row having first row address values. A mask has mask bits corresponding to each of the first row address values. The mask bits are initially in active status. A failure of a second row, having second row address values, is detected. When none of the first row address values matches the second row address values, and when mask bits are all in the active status, the array macros are determined to be bad. When at least one of the first row address values matches the second row address values, mask bits that correspond to at least one of the first row address values that match are kept in active status, and mask bits that correspond to non-matching first address values are set to inactive status.

Address translation of host commands to access host data stored in memory devices that provides a chip kill capability not only involves locating where the host data is stored, but also involves locating where parity data striped with the host data is stored. In locating where the parity data is stored, the address translation can be performed with logical (e.g., arithmetic) operations.