Hi jess,
Thanks for your soon answer. I forgot to tell That I have only ONE SPHINXS SERVER. Because to solve this problem i have reduce all my enviroment to one node for each service.
I Have check the logs /opt/kaltura/log/sphinx/kaltura_sphinx_searchd.log… and you were right I have the error
"FATAL: failed to lock pid file ‘/opt/kaltura/sphinx/searchd.pid’: Resource temporarily unavailable (searchd already running?)" So I think That I have one proccess already running… So how can I stop this zombie proccess, reindex and restart the index correctly… Please I wolud like the correct steps… Maybe I am doing wrong beacuse I am changing the steps order