Solaris Dynamic Tracing Guide

Speculation Options and Tuning

If a speculative buffer is full when a speculative tracing action is attempted, no data is stored in the buffer and a drop count is incremented. If this situation, a dtrace message similar to the following example is generated:


dtrace: 38 speculative drops

Speculative drops will not prevent the full speculative buffer from being copied into the principal buffer when the buffer is committed. Similarly, speculative drops can occur even if drops were experienced on a speculative buffer that was ultimately discarded. Speculative drops can be reduced by increasing the speculative buffer size, which is tuned using the specsize option. The specsize option may be specified with any size suffix. The resizing policy of this buffer is dictated by the bufresize option.

Speculative buffers might be unavailable when speculation() is called. If buffers exist that have not yet been committed or discarded, a dtrace message similar to the following example is generated:


dtrace: 1 failed speculation (no speculative buffer available)

You can reduce the likelihood of failed speculations of this nature by increasing the number of speculative buffers with the nspec option. The value of nspec defaults to one.

Alternatively, speculation() may fail because all speculative buffers are busy. In this case, a dtrace message similar to the following example is generated:


dtrace: 1 failed speculation (available buffer(s) still busy)

This message indicates that speculation() was called after commit() was called for a speculative buffer, but before that buffer was actually committed on all CPUs. You can reduce the likelihood of failed speculations of this nature by increasing the rate at which CPUs are cleaned with the cleanrate option. The value of cleanrate defaults to 101hz.


Note –

You must specify values for the cleanrate option in number-per-second. Use the hz suffix.