В книге "Training Kit (Exam 70-461): Querying Microsoft SQL Server 2012"
Есть такой абзац:
Starting with sequential keys, all rows go into the right end of the index. When a page is
full, SQL Server allocates a new page and fills it. This results in less fragmentation in the index,
which is beneficial for read performance. Also, insertions can be faster when a single session
is loading the data, and the data resides on a single drive or a small number of drives.
However, with high-end storage subsystems that have many spindles, the situation can be
different. When loading data from multiple sessions, you will end up with page latch contention
(latches are objects used to synchronize access to database pages) against the rightmost
pages of the index leaf level’s linked list. This bottleneck prevents use of the full throughput
of the storage subsystem.
Подскажите, что здесь подразумевается под spindles? что значит "high-end storage subsystems that have many spindles"?
Есть такой абзац:
Starting with sequential keys, all rows go into the right end of the index. When a page is
full, SQL Server allocates a new page and fills it. This results in less fragmentation in the index,
which is beneficial for read performance. Also, insertions can be faster when a single session
is loading the data, and the data resides on a single drive or a small number of drives.
However, with high-end storage subsystems that have many spindles, the situation can be
different. When loading data from multiple sessions, you will end up with page latch contention
(latches are objects used to synchronize access to database pages) against the rightmost
pages of the index leaf level’s linked list. This bottleneck prevents use of the full throughput
of the storage subsystem.
Подскажите, что здесь подразумевается под spindles? что значит "high-end storage subsystems that have many spindles"?