Hi,
We've seen this happen a couple of times before. We haven't been able to
identify a likely suspect yet. What we think is happening is that a
non-Accumulo process happens to connect to an Accumulo process, send a
message that isn't a Thrift message (that Accumulo uses internally) and
it happens to try to allocate an extremely large buffer.
https://issues.apache.org/jira/browse/ACCUMULO-2360
We introduced an upper limit on the size of an RPC message and this is
what this error message is telling you. Are you saying that this is
still causing the Master process to crash? If so, are there any error
messages at the end of the .log, .out or .err files?
[email protected] wrote:
HI,
I recently learned Accumulo,but it master server always abnormal,then
not runing.
hadoop.2.5.1 zookeeper 3.4.6 accumulo 1.6.0/1.6.1
[util.TServerUtils$THsHaServer] ERROR: Read a frame size of 1195725856, which
is bigger than the maximum allowable buffer size for ALL connections.
2014-11-20 15:02:35,974 [util.TServerUtils$THsHaServer] ERROR: Read a frame
size of 1330664521, which is bigger than the maximum allowable buffer size for
ALL connections.
2014-11-20 15:02:35,977 [util.TServerUtils$THsHaServer] ERROR: Read a frame
size of 1195725856, which is bigger than the maximum allowable buffer size for
ALL connections.
[email protected]
