Hi,

I have a question about TTS efficiency. I have some text that I need to read out when someone calls in and these texts do change. Here are two options I am thinking about:
In my configuration, I would have DB and FS on different servers. DB server is SCSI.

Option 1:Store the text in DB

Within JS:
1.Get the text from DB
2. use system (command) to create a wav file on locate file system
3. Use playback to play the wav file

Option 2: Pre-created the file using NFS ( at the same machine as DB)
Within JS:
1. Get the filename from DB
2. Get the wav file from NFS
3. Use playback to play the wav file

In a highly-loaded environment, which option would be preferable considering speed and efficiency?

Is there other better option?

Any input will be greatly appreciated.

Thanks,
Pete