Hello Sailesh Mukil,
I'd like you to reexamine a change. Please visit
http://gerrit.cloudera.org:8080/3343
to look at the new patch set (#14).
Change subject: IMPALA-3575: Add retry to backend connection request and rpc
timeout
......................................................................
IMPALA-3575: Add retry to backend connection request and rpc timeout
This patch adds a configurable timeout for all backend client
RPC calls to avoid query hang issue.
Impala doesn't set socket send/recv timeout for backend client.
RPC calls will wait forever for data. In extreme case of bad network,
or destination host has kernel panic, sender will not get response
and rpc call will hang. Query hang is hard to detect. if hang happens
at ExecRemoteFragment() or CancelPlanFragments(), query cannot be
canelled unless you restart coordinator.
Added send/recv timeout to all rpc calls to avoid query hang. And fix
a bug that reporting thread does not quiting even after query is cancelled.
For catalog client, keep default timeout to 0 (no timeout) because ExecDdl()
could take very long time if table has many partitons, mainly waiting for
HMS API call.
Added a new RPC call DoRpcTimedWait() to wait for receiver response for
longer time. This is needed by certain RPCs. For example, TransmitData()
by DataStreamSender, receiver could hold response due to back pressure.
If an RPC call fails, we don't put the underlying connection back to
cache but close it. This is to make sure bad state of this connection
won't cause more RPC failure.
Added retry for CancelPlanFragment RPC. This reduces the chance that cancel
request gets lost due to unstable network, but this can cause cancellation
takes longer time. and make test_lifecycle.py more flaky.
The metric num-fragments-in-flight might not be 0 yet due to previous tests.
Modified the test to check the metric delta instead of comparing to 0 to
reduce flakyness. However, this might not capture some failures.
Besides the new EE test, I used the following iptable rule to
inject network failure to make sure rpc call never hang.
1. Block network traffic on a port completely
iptables -A INPUT -p tcp -m tcp --dport 22002 -j DROP
2. Randomly drop 5% of TCP packet to slowdown network
iptables -A INPUT -p tcp -m tcp --dport 22000 -m statistic --mode random
--probability 0.05 -j DROP
Change-Id: Id6723cfe58df6217f4a9cdd12facd320cbc24964
---
M be/src/runtime/client-cache.cc
M be/src/runtime/client-cache.h
M be/src/runtime/coordinator.cc
M be/src/runtime/data-stream-sender.cc
M be/src/runtime/exec-env.cc
M be/src/runtime/plan-fragment-executor.cc
M be/src/service/fragment-exec-state.cc
M be/src/statestore/statestore.cc
M common/thrift/generate_error_codes.py
A tests/custom_cluster/test_rpc_timeout.py
M tests/query_test/test_lifecycle.py
M tests/verifiers/metric_verifier.py
12 files changed, 222 insertions(+), 32 deletions(-)
git pull ssh://gerrit.cloudera.org:29418/Impala refs/changes/43/3343/14
--
To view, visit http://gerrit.cloudera.org:8080/3343
To unsubscribe, visit http://gerrit.cloudera.org:8080/settings
Gerrit-MessageType: newpatchset
Gerrit-Change-Id: Id6723cfe58df6217f4a9cdd12facd320cbc24964
Gerrit-PatchSet: 14
Gerrit-Project: Impala
Gerrit-Branch: cdh5-trunk
Gerrit-Owner: Juan Yu <j...@cloudera.com>
Gerrit-Reviewer: Alan Choi <a...@cloudera.com>
Gerrit-Reviewer: Dan Hecht <dhe...@cloudera.com>
Gerrit-Reviewer: Henry Robinson <he...@cloudera.com>
Gerrit-Reviewer: Juan Yu <j...@cloudera.com>
Gerrit-Reviewer: Sailesh Mukil <sail...@cloudera.com>
