[jira] [Commented] (CASSANDRA-11720) Changing `max_hint_window_in_ms` at runtime
[ https://issues.apache.org/jira/browse/CASSANDRA-11720?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16002485#comment-16002485 ] Hiroyuki Nishi commented on CASSANDRA-11720: [~michaelsembwever] Looks good. I appreciate your response to update the patch and add the dtest! > Changing `max_hint_window_in_ms` at runtime > --- > > Key: CASSANDRA-11720 > URL: https://issues.apache.org/jira/browse/CASSANDRA-11720 > Project: Cassandra > Issue Type: Wish > Components: Coordination >Reporter: Jens Rantil >Assignee: Hiroyuki Nishi >Priority: Minor > Labels: lhf > Fix For: 4.x > > Attachments: CASSANDRA-11720-trunk.patch > > > Scenario: A larger node (in terms of data it holds) goes down. You realize > that it will take slightly more than `max_hint_window_in_ms` to fix it. You > have a the disk space to store some additional hints. > Proposal: Support changing `max_hint_window_in_ms` at runtime. The change > doesn't have to be persisted somewhere. I'm thinking similar to changing the > `compactionthroughput` etc. using `nodetool`. > Workaround: Change the value in the configuration file and do a rolling > restart of all the nodes. -- This message was sent by Atlassian JIRA (v6.3.15#6346) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-11720) Changing `max_hint_window_in_ms` at runtime
[ https://issues.apache.org/jira/browse/CASSANDRA-11720?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15994171#comment-15994171 ] mck commented on CASSANDRA-11720: - Patch ready. || Branch || Testall || Dtest || | [trunk|https://github.com/michaelsembwever/cassandra/commit/f2344741a560c44538e88fb02cffa87f74445c9d] | [testall|https://circleci.com/gh/michaelsembwever/cassandra/tree/mck%2Ftrunk_11720] | [dtest|https://builds.apache.org/view/A-D/view/Cassandra/job/Cassandra-devbranch-dtest/39/] | (the dtests will take a while, and won't include the new [dtest|https://github.com/michaelsembwever/cassandra-dtest/commit/f5267617cb23b8faa968ca2a11ccf49d0725a81d]) > Changing `max_hint_window_in_ms` at runtime > --- > > Key: CASSANDRA-11720 > URL: https://issues.apache.org/jira/browse/CASSANDRA-11720 > Project: Cassandra > Issue Type: Wish > Components: Coordination >Reporter: Jens Rantil >Assignee: Hiroyuki Nishi >Priority: Minor > Labels: lhf > Fix For: 4.x > > Attachments: CASSANDRA-11720-trunk.patch > > > Scenario: A larger node (in terms of data it holds) goes down. You realize > that it will take slightly more than `max_hint_window_in_ms` to fix it. You > have a the disk space to store some additional hints. > Proposal: Support changing `max_hint_window_in_ms` at runtime. The change > doesn't have to be persisted somewhere. I'm thinking similar to changing the > `compactionthroughput` etc. using `nodetool`. > Workaround: Change the value in the configuration file and do a rolling > restart of all the nodes. -- This message was sent by Atlassian JIRA (v6.3.15#6346) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-11720) Changing `max_hint_window_in_ms` at runtime
[ https://issues.apache.org/jira/browse/CASSANDRA-11720?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15994168#comment-15994168 ] mck commented on CASSANDRA-11720: - [~hnishi], my patch is updated with your modified description. that patch is now at https://github.com/michaelsembwever/cassandra/commit/f2344741a560c44538e88fb02cffa87f74445c9d And the dtest patch is here – https://github.com/michaelsembwever/cassandra-dtest/commit/f5267617cb23b8faa968ca2a11ccf49d0725a81d > Changing `max_hint_window_in_ms` at runtime > --- > > Key: CASSANDRA-11720 > URL: https://issues.apache.org/jira/browse/CASSANDRA-11720 > Project: Cassandra > Issue Type: Wish > Components: Coordination >Reporter: Jens Rantil >Assignee: Hiroyuki Nishi >Priority: Minor > Labels: lhf > Fix For: 4.x > > Attachments: CASSANDRA-11720-trunk.patch > > > Scenario: A larger node (in terms of data it holds) goes down. You realize > that it will take slightly more than `max_hint_window_in_ms` to fix it. You > have a the disk space to store some additional hints. > Proposal: Support changing `max_hint_window_in_ms` at runtime. The change > doesn't have to be persisted somewhere. I'm thinking similar to changing the > `compactionthroughput` etc. using `nodetool`. > Workaround: Change the value in the configuration file and do a rolling > restart of all the nodes. -- This message was sent by Atlassian JIRA (v6.3.15#6346) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-11720) Changing `max_hint_window_in_ms` at runtime
[ https://issues.apache.org/jira/browse/CASSANDRA-11720?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15994151#comment-15994151 ] Hiroyuki Nishi commented on CASSANDRA-11720: [~michaelsembwever] I'm sorry that you are confused in my English. Thank you for updating the patch. I would like to modify the description of setmaxhintwindow additionally as follows. https://github.com/yhnishi/cassandra/commit/437c659f13ad430c3cff260aa94ea3f706e464f5 Please let me know what you think. > Changing `max_hint_window_in_ms` at runtime > --- > > Key: CASSANDRA-11720 > URL: https://issues.apache.org/jira/browse/CASSANDRA-11720 > Project: Cassandra > Issue Type: Wish > Components: Coordination >Reporter: Jens Rantil >Assignee: Hiroyuki Nishi >Priority: Minor > Labels: lhf > Fix For: 4.x > > Attachments: CASSANDRA-11720-trunk.patch > > > Scenario: A larger node (in terms of data it holds) goes down. You realize > that it will take slightly more than `max_hint_window_in_ms` to fix it. You > have a the disk space to store some additional hints. > Proposal: Support changing `max_hint_window_in_ms` at runtime. The change > doesn't have to be persisted somewhere. I'm thinking similar to changing the > `compactionthroughput` etc. using `nodetool`. > Workaround: Change the value in the configuration file and do a rolling > restart of all the nodes. -- This message was sent by Atlassian JIRA (v6.3.15#6346) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-11720) Changing `max_hint_window_in_ms` at runtime
[ https://issues.apache.org/jira/browse/CASSANDRA-11720?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15993703#comment-15993703 ] mck commented on CASSANDRA-11720: - {quote} I think that i will implement the above correspondence, how is it?{quote} I'm not sure I understand you [~hnishi]. I've updated your patch, it is here – https://github.com/michaelsembwever/cassandra/commit/b40cb59728112b759b6caa8204152add825829cd I'm looking into adding something in to the existing hints dtest found in [hintedhandoff_test.py|https://github.com/riptano/cassandra-dtest/blob/master/hintedhandoff_test.py] > Changing `max_hint_window_in_ms` at runtime > --- > > Key: CASSANDRA-11720 > URL: https://issues.apache.org/jira/browse/CASSANDRA-11720 > Project: Cassandra > Issue Type: Wish > Components: Coordination >Reporter: Jens Rantil >Assignee: Hiroyuki Nishi >Priority: Minor > Labels: lhf > Fix For: 4.x > > Attachments: CASSANDRA-11720-trunk.patch > > > Scenario: A larger node (in terms of data it holds) goes down. You realize > that it will take slightly more than `max_hint_window_in_ms` to fix it. You > have a the disk space to store some additional hints. > Proposal: Support changing `max_hint_window_in_ms` at runtime. The change > doesn't have to be persisted somewhere. I'm thinking similar to changing the > `compactionthroughput` etc. using `nodetool`. > Workaround: Change the value in the configuration file and do a rolling > restart of all the nodes. -- This message was sent by Atlassian JIRA (v6.3.15#6346) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-11720) Changing `max_hint_window_in_ms` at runtime
[ https://issues.apache.org/jira/browse/CASSANDRA-11720?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15992558#comment-15992558 ] Hiroyuki Nishi commented on CASSANDRA-11720: [~michaelsembwever] I am very sorry for the late reply... I think that i will implement the above correspondence, how is it? > Changing `max_hint_window_in_ms` at runtime > --- > > Key: CASSANDRA-11720 > URL: https://issues.apache.org/jira/browse/CASSANDRA-11720 > Project: Cassandra > Issue Type: Wish > Components: Coordination >Reporter: Jens Rantil >Assignee: Hiroyuki Nishi >Priority: Minor > Labels: lhf > Fix For: 4.x > > Attachments: CASSANDRA-11720-trunk.patch > > > Scenario: A larger node (in terms of data it holds) goes down. You realize > that it will take slightly more than `max_hint_window_in_ms` to fix it. You > have a the disk space to store some additional hints. > Proposal: Support changing `max_hint_window_in_ms` at runtime. The change > doesn't have to be persisted somewhere. I'm thinking similar to changing the > `compactionthroughput` etc. using `nodetool`. > Workaround: Change the value in the configuration file and do a rolling > restart of all the nodes. -- This message was sent by Atlassian JIRA (v6.3.15#6346) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-11720) Changing `max_hint_window_in_ms` at runtime
[ https://issues.apache.org/jira/browse/CASSANDRA-11720?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15992532#comment-15992532 ] mck commented on CASSANDRA-11720: - thanks [~rha]! i'll update the patch. i'm also looking into adding something to the {{hintedhandoff_test.py}} dtest. > Changing `max_hint_window_in_ms` at runtime > --- > > Key: CASSANDRA-11720 > URL: https://issues.apache.org/jira/browse/CASSANDRA-11720 > Project: Cassandra > Issue Type: Wish > Components: Coordination >Reporter: Jens Rantil >Assignee: Hiroyuki Nishi >Priority: Minor > Labels: lhf > Fix For: 4.x > > Attachments: CASSANDRA-11720-trunk.patch > > > Scenario: A larger node (in terms of data it holds) goes down. You realize > that it will take slightly more than `max_hint_window_in_ms` to fix it. You > have a the disk space to store some additional hints. > Proposal: Support changing `max_hint_window_in_ms` at runtime. The change > doesn't have to be persisted somewhere. I'm thinking similar to changing the > `compactionthroughput` etc. using `nodetool`. > Workaround: Change the value in the configuration file and do a rolling > restart of all the nodes. -- This message was sent by Atlassian JIRA (v6.3.15#6346) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-11720) Changing `max_hint_window_in_ms` at runtime
[ https://issues.apache.org/jira/browse/CASSANDRA-11720?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15992489#comment-15992489 ] Romain Hardouin commented on CASSANDRA-11720: - It would be nice to have that in nodetool. I use jmxterm to do that currently: {code} echo "set -b org.apache.cassandra.db:type=StorageProxy MaxHintWindow " | java -jar /path/to/jmxterm.jar -l 127.0.0.1:7199 -u ... -p ... {code} Some comments: In {{SetMaxHintWindow}} it should be {{value_in_ms}}: {code} usage = "" {code} In {{GetMaxHintWindow}} I don't understand "of the given type" here: {code} @Command(name = "getmaxhintwindow", description = "Print the max hint window of the given type in ms") {code} Maybe just {{Print max hint window in ms}}? Thanks > Changing `max_hint_window_in_ms` at runtime > --- > > Key: CASSANDRA-11720 > URL: https://issues.apache.org/jira/browse/CASSANDRA-11720 > Project: Cassandra > Issue Type: Wish > Components: Coordination >Reporter: Jens Rantil >Assignee: Hiroyuki Nishi >Priority: Minor > Labels: lhf > Fix For: 4.x > > Attachments: CASSANDRA-11720-trunk.patch > > > Scenario: A larger node (in terms of data it holds) goes down. You realize > that it will take slightly more than `max_hint_window_in_ms` to fix it. You > have a the disk space to store some additional hints. > Proposal: Support changing `max_hint_window_in_ms` at runtime. The change > doesn't have to be persisted somewhere. I'm thinking similar to changing the > `compactionthroughput` etc. using `nodetool`. > Workaround: Change the value in the configuration file and do a rolling > restart of all the nodes. -- This message was sent by Atlassian JIRA (v6.3.15#6346) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-11720) Changing `max_hint_window_in_ms` at runtime
[ https://issues.apache.org/jira/browse/CASSANDRA-11720?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15991718#comment-15991718 ] mck commented on CASSANDRA-11720: - [~hnishi], did you get a chance to look at this? > Changing `max_hint_window_in_ms` at runtime > --- > > Key: CASSANDRA-11720 > URL: https://issues.apache.org/jira/browse/CASSANDRA-11720 > Project: Cassandra > Issue Type: Wish > Components: Coordination >Reporter: Jens Rantil >Assignee: Hiroyuki Nishi >Priority: Minor > Labels: lhf > Fix For: 3.11.x > > Attachments: CASSANDRA-11720-trunk.patch > > > Scenario: A larger node (in terms of data it holds) goes down. You realize > that it will take slightly more than `max_hint_window_in_ms` to fix it. You > have a the disk space to store some additional hints. > Proposal: Support changing `max_hint_window_in_ms` at runtime. The change > doesn't have to be persisted somewhere. I'm thinking similar to changing the > `compactionthroughput` etc. using `nodetool`. > Workaround: Change the value in the configuration file and do a rolling > restart of all the nodes. -- This message was sent by Atlassian JIRA (v6.3.15#6346) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-11720) Changing `max_hint_window_in_ms` at runtime
[ https://issues.apache.org/jira/browse/CASSANDRA-11720?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15649054#comment-15649054 ] mck commented on CASSANDRA-11720: - LGTM. Some small feedback [~ztyx]: • can we add a test? ∘ dtest (nodetool_test.py) (or a unit test if something can be created…) • the docs need to be updated ∘ we'd need an entry in {{doc/build/html/tools/nodetool/nodetool.html}} > Changing `max_hint_window_in_ms` at runtime > --- > > Key: CASSANDRA-11720 > URL: https://issues.apache.org/jira/browse/CASSANDRA-11720 > Project: Cassandra > Issue Type: Wish > Components: Coordination >Reporter: Jens Rantil >Priority: Minor > Labels: lhf > Attachments: CASSANDRA-11720-trunk.patch > > > Scenario: A larger node (in terms of data it holds) goes down. You realize > that it will take slightly more than `max_hint_window_in_ms` to fix it. You > have a the disk space to store some additional hints. > Proposal: Support changing `max_hint_window_in_ms` at runtime. The change > doesn't have to be persisted somewhere. I'm thinking similar to changing the > `compactionthroughput` etc. using `nodetool`. > Workaround: Change the value in the configuration file and do a rolling > restart of all the nodes. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (CASSANDRA-11720) Changing `max_hint_window_in_ms` at runtime
[ https://issues.apache.org/jira/browse/CASSANDRA-11720?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15490546#comment-15490546 ] Hiroyuki Nishi commented on CASSANDRA-11720: Hello, I fixed this issue. Below is a sample result. {code} $ ./bin/nodetool getmaxhintwindow Current max hint window: 1080 ms $ ./bin/nodetool setmaxhintwindow 0 $ ./bin/nodetool getmaxhintwindow Current max hint window: 0 ms {code} Could someone review the attached patch? I worked in trunk branch. Thank you. > Changing `max_hint_window_in_ms` at runtime > --- > > Key: CASSANDRA-11720 > URL: https://issues.apache.org/jira/browse/CASSANDRA-11720 > Project: Cassandra > Issue Type: Wish > Components: Coordination >Reporter: Jens Rantil >Priority: Minor > Labels: lhf > Attachments: CASSANDRA-11720-trunk.patch > > > Scenario: A larger node (in terms of data it holds) goes down. You realize > that it will take slightly more than `max_hint_window_in_ms` to fix it. You > have a the disk space to store some additional hints. > Proposal: Support changing `max_hint_window_in_ms` at runtime. The change > doesn't have to be persisted somewhere. I'm thinking similar to changing the > `compactionthroughput` etc. using `nodetool`. > Workaround: Change the value in the configuration file and do a rolling > restart of all the nodes. -- This message was sent by Atlassian JIRA (v6.3.4#6332)