- Receive request failed. Err: Estream construct failed. Err: [STREAM] Bad input stream Err: [STREAM] Bad input stream - no info
- Agent request time out Err:Estream construct failed. Err: RWSockErr: in RWSocket::recv:INTR - no info
From EMC Powerlink site:
These errors are caused by latency in the network. They are benign and can be ignored. They are an indication that the host Agent did not receive a response from the array. A request was sent but no response was received and the command timed out. Changing the poll interval in the Agent can sometimes help by reducing contention for Navisphere agent resources on the array, but more usually it turns out to a network issue. There may be periods when network traffic is higher than usual, and responses to requests are delayed. It may also be noticed that hosts occasionally appear as "unmanaged, " then as "managed."
This is a Navisphere Agent message that should only occur on HP-UX. This happens because the Agent started sending some data, but did not finish before a Navisphere CLI command hit a timeout. It appears as "INTR" because HP-UX is not compiled with POSIX thread support, and thus calls alarm() which. in turn, causes a Rogue Wave INTR exception. (Rogue Wave is a cross platform software tool.)
Fix: Investigate the possibility of latency in the network.
[Note from user: I am getting these on a Solaris 5.8 machine connected to a Clariion FC4700, so they are not unique to HP-UX.]