You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@hbase.apache.org by st...@apache.org on 2021/03/16 04:33:03 UTC
[hbase] branch branch-2.4 updated: Revert "HBASE-25663 Make
graceful_stop localhostname compare match even if fqdn (#3048)"
This is an automated email from the ASF dual-hosted git repository.
stack pushed a commit to branch branch-2.4
in repository https://gitbox.apache.org/repos/asf/hbase.git
The following commit(s) were added to refs/heads/branch-2.4 by this push:
new 3aa7285 Revert "HBASE-25663 Make graceful_stop localhostname compare match even if fqdn (#3048)"
3aa7285 is described below
commit 3aa72858c929a098ffc158de2407a510106507ad
Author: stack <st...@apache.org>
AuthorDate: Mon Mar 15 21:32:28 2021 -0700
Revert "HBASE-25663 Make graceful_stop localhostname compare match even if fqdn (#3048)"
This reverts commit 5366a7c4b01fe39ecd80717143b9fcd52fb57666.
---
bin/graceful_stop.sh | 52 ++++++++++++++++++++++++++------
src/main/asciidoc/_chapters/ops_mgt.adoc | 22 +++++++-------
2 files changed, 53 insertions(+), 21 deletions(-)
diff --git a/bin/graceful_stop.sh b/bin/graceful_stop.sh
index ac5b641..a2b4efe 100755
--- a/bin/graceful_stop.sh
+++ b/bin/graceful_stop.sh
@@ -18,7 +18,8 @@
# * limitations under the License.
# */
-# Move regions off a server then stop it. Optionally restart and reload.
+# Move regions off a server then stop it. Optionally restart and reload.
+# Turn off the balancer before running this script.
function usage {
echo "Usage: graceful_stop.sh [--config <conf-dir>] [-e] [--restart [--reload]] [--thrift] \
[--rest] [-n |--noack] [--maxthreads <number of threads>] [--movetimeout <timeout in seconds>] \
@@ -32,7 +33,7 @@ moving regions"
echo " maxthreads xx Limit the number of threads used by the region mover. Default value is 1."
echo " movetimeout xx Timeout for moving regions. If regions are not moved by the timeout value,\
exit with error. Default value is INT_MAX."
- echo " hostname Hostname to stop; match what HBase uses; pass 'localhost' if local to avoid ssh"
+ echo " hostname Hostname of server we are to stop"
echo " e|failfast Set -e so exit immediately if any command exits with non-zero status"
echo " nob|nobalancer Do not manage balancer states. This is only used as optimization in \
rolling_restart.sh to avoid multiple calls to hbase shell"
@@ -101,6 +102,13 @@ fi
hostname=$1
filename="/tmp/$hostname"
+local=
+localhostname=`/bin/hostname`
+
+if [ "$localhostname" == "$hostname" ]; then
+ local=true
+fi
+
if [ "$nob" == "true" ]; then
log "[ $0 ] skipping disabling balancer -nob argument is used"
HBASE_BALANCER_STATE=false
@@ -111,7 +119,7 @@ else
fi
unload_args="--filename $filename --maxthreads $maxthreads $noack --operation unload \
- --timeout $movetimeout --regionserverhost $hostname"
+--timeout $movetimeout --regionserverhost $hostname"
if [ "$designatedfile" != "" ]; then
unload_args="$unload_args --designatedfile $designatedfile"
@@ -131,25 +139,49 @@ hosts="/tmp/$(basename $0).$$.tmp"
echo $hostname >> $hosts
if [ "$thrift" != "" ]; then
log "Stopping thrift server on $hostname"
- "$bin"/hbase-daemons.sh --config ${HBASE_CONF_DIR} --hosts ${hosts} stop thrift
+ if [ "$local" == true ]; then
+ "$bin"/hbase-daemon.sh --config ${HBASE_CONF_DIR} stop thrift
+ else
+ "$bin"/hbase-daemons.sh --config ${HBASE_CONF_DIR} --hosts ${hosts} stop thrift
+ fi
fi
if [ "$rest" != "" ]; then
log "Stopping rest server on $hostname"
- "$bin"/hbase-daemons.sh --config ${HBASE_CONF_DIR} --hosts ${hosts} stop rest
+ if [ "$local" == true ]; then
+ "$bin"/hbase-daemon.sh --config ${HBASE_CONF_DIR} stop rest
+ else
+ "$bin"/hbase-daemons.sh --config ${HBASE_CONF_DIR} --hosts ${hosts} stop rest
+ fi
fi
log "Stopping regionserver on $hostname"
-"$bin"/hbase-daemons.sh --config ${HBASE_CONF_DIR} --hosts ${hosts} stop regionserver
+if [ "$local" == true ]; then
+ "$bin"/hbase-daemon.sh --config ${HBASE_CONF_DIR} stop regionserver
+else
+ "$bin"/hbase-daemons.sh --config ${HBASE_CONF_DIR} --hosts ${hosts} stop regionserver
+fi
if [ "$restart" != "" ]; then
log "Restarting regionserver on $hostname"
- "$bin"/hbase-daemons.sh --config ${HBASE_CONF_DIR} --hosts ${hosts} start regionserver
+ if [ "$local" == true ]; then
+ "$bin"/hbase-daemon.sh --config ${HBASE_CONF_DIR} start regionserver
+ else
+ "$bin"/hbase-daemons.sh --config ${HBASE_CONF_DIR} --hosts ${hosts} start regionserver
+ fi
if [ "$thrift" != "" ]; then
log "Restarting thrift server on $hostname"
# -b 0.0.0.0 says listen on all interfaces rather than just default.
- "$bin"/hbase-daemons.sh --config ${HBASE_CONF_DIR} --hosts ${hosts} start thrift -b 0.0.0.0
+ if [ "$local" == true ]; then
+ "$bin"/hbase-daemon.sh --config ${HBASE_CONF_DIR} start thrift -b 0.0.0.0
+ else
+ "$bin"/hbase-daemons.sh --config ${HBASE_CONF_DIR} --hosts ${hosts} start thrift -b 0.0.0.0
+ fi
fi
if [ "$rest" != "" ]; then
log "Restarting rest server on $hostname"
- "$bin"/hbase-daemons.sh --config ${HBASE_CONF_DIR} --hosts ${hosts} start rest
+ if [ "$local" == true ]; then
+ "$bin"/hbase-daemon.sh --config ${HBASE_CONF_DIR} start rest
+ else
+ "$bin"/hbase-daemons.sh --config ${HBASE_CONF_DIR} --hosts ${hosts} start rest
+ fi
fi
if [ "$reload" != "" ]; then
log "Reloading $hostname region(s)"
@@ -169,4 +201,4 @@ else
fi
# Cleanup tmp files.
-trap "rm -f /tmp/$(basename $0).*.tmp &> /dev/null" EXIT
+trap "rm -f "/tmp/$(basename $0).*.tmp" &> /dev/null" EXIT
diff --git a/src/main/asciidoc/_chapters/ops_mgt.adoc b/src/main/asciidoc/_chapters/ops_mgt.adoc
index a79f14a..d1faa60 100644
--- a/src/main/asciidoc/_chapters/ops_mgt.adoc
+++ b/src/main/asciidoc/_chapters/ops_mgt.adoc
@@ -1364,9 +1364,10 @@ Copy the script if you need to make use of it in a version of hbase previous to
A downside to the above stop of a RegionServer is that regions could be offline for a good period of time.
Regions are closed in order.
-If many regions on the server, the first region to close may not be back online until all regions close and
-after the master notices the RegionServer's znode gone. A node can be asked to gradually shed its load and
-then shutdown itself using the _graceful_stop.sh_ script. Here is its usage:
+If many regions on the server, the first region to close may not be back online until all regions close and after the master notices the RegionServer's znode gone.
+In Apache HBase 0.90.2, we added facility for having a node gradually shed its load and then shutdown itself down.
+Apache HBase 0.90.2 added the _graceful_stop.sh_ script.
+Here is its usage:
----
$ ./bin/graceful_stop.sh
@@ -1392,17 +1393,16 @@ To decommission a loaded RegionServer, run the following: +$
[NOTE]
====
The `HOSTNAME` passed to _graceful_stop.sh_ must match the hostname that hbase is using to identify RegionServers.
-HBase uses fully-qualified domain names usually. Check the list of RegionServers in the master UI for how HBase
-is referring to servers. Whatever HBase is using, this is what you should pass the _graceful_stop.sh_ decommission script.
-If you pass IPs, the script is not yet smart enough to make a hostname (or FQDN) of it and so it will fail when it checks
-if server is currently running; the graceful unloading of regions will not run.
+Check the list of RegionServers in the master UI for how HBase is referring to servers.
+It's usually hostname but can also be FQDN.
+Whatever HBase is using, this is what you should pass the _graceful_stop.sh_ decommission script.
+If you pass IPs, the script is not yet smart enough to make a hostname (or FQDN) of it and so it will fail when it checks if server is currently running; the graceful unloading of regions will not run.
====
The _graceful_stop.sh_ script will move the regions off the decommissioned RegionServer one at a time to minimize region churn.
-It will verify the region deployed in the new location before it will moves the next region and so on until the decommissioned
-server is carrying zero regions. At this point, the _graceful_stop.sh_ tells the RegionServer `stop`.
-The master will at this point notice the RegionServer gone but all regions will have already been redeployed and because the
-RegionServer went down cleanly, there will be no WAL logs to split.
+It will verify the region deployed in the new location before it will moves the next region and so on until the decommissioned server is carrying zero regions.
+At this point, the _graceful_stop.sh_ tells the RegionServer `stop`.
+The master will at this point notice the RegionServer gone but all regions will have already been redeployed and because the RegionServer went down cleanly, there will be no WAL logs to split.
[[lb]]
.Load Balancer