Request timeout errors logged by offset mananger

[horkhe]

Some hosts are generating a bunch of request timeout errors e.g.:

</cons[0]/G:integrations[0]/P:account_events_1[0]/offset_mgr[0]> trigger reassign: reason=offset commit failed, err=(request timeout)

Although it does not seem to cause neither message loses nor duplicate consumptions.