On 11/23/12 10:13 AM, ming....@gmail.com wrote:
'why mt ats is crashing', I have to tell them: 1, do not enable Range in v3.2.0
Right, fixed on trunk I assume / hope?
2, do not use the CLFUS, use the standard LRU
What is the issue here? Are there bugs filed for this? If CLFUS is crashing, we should fix it (and/or perhaps change the default).
3, do not use the thread pool to manage server sessions
If I recall, this is something with the cluster interaction with the sessions? What are the bug ID's here? For most users, the above is bad advice IMO. If you don't use cache clustering, the per thread pools of server sessions is vastly superior (by 1 magnitude or more).
Cheers, -- leif
