Gus, we've found the cause.

It was a problem in Cassandra, but it has been already fixed in cassandra 1.1.6.

Commit with the problem:
2c69e2ea757be9492a095aa22b5d51234c4b4102
You can see it at https://issues.apache.org/jira/secure/attachment/12544204/CASSANDRA-4561-CS.patch

Commit with the fix:
988ea81d409968614d84dacb3a022dcb156172c3
There is no ticket in JIRA about that commit (at least I couldn't find the ticket). 

Also our client node just was not synchronized accordingly Cassandra's nodes. Client node lived "in the future" (just a few minutes). 

So that's the cause of described streams during repair process.

Thanks all for the discussion!