It's more likely a software issue in the protocol stack that negotiates and control the switching fabric.
I work in telecoms and specialized in development and testing of TDM for almost a decade. I never saw clock slip manifest itself as a whole channel shift. The drift occurs at bit level, not in chunks of 8 bits. The elastic buffers in the phy protect against behavior and if they overflow due to a persistent drift it only results in a click in the channel.
I work in telecoms and specialized in development and testing of TDM for almost a decade. I never saw clock slip manifest itself as a whole channel shift. The drift occurs at bit level, not in chunks of 8 bits. The elastic buffers in the phy protect against behavior and if they overflow due to a persistent drift it only results in a click in the channel.