[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17736238#comment-17736238 ] Brad Schoening commented on CASSANDRA-18305: Per hour is fine. I had considered suggesting that also. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 3h 10m > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17736236#comment-17736236 ] Stefan Miklosovic commented on CASSANDRA-18305: --- [~bschoeni] would you mind if mean rate would be {_}per hour{_}? If you prefer it per minute, I still want to have a possibility to have it per hour without multiplying it by 60 again in my head. We may have a flag for it. If mean rate per hour is OK for you we can do it like that and drop the flag. Cassandra is a process which is meant to be run non-stop and if an operator wants to see how are compactions going, I doubt the metric "per minute" for mean rate is useful. I think that extrapolating that number to represent mean number of compactions per hour is more telling. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 3h 10m > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17736219#comment-17736219 ] Brad Schoening commented on CASSANDRA-18305: [~smiklosovic] yes, the above would be useful IMHO and it does tie out directly with pending compactions – assuming a task means sstable. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 3h 10m > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17736217#comment-17736217 ] Brandon Williams commented on CASSANDRA-18305: -- bq. why was it difficult to have compaction throughput (Mbps) as a JMX Meter? Guava's RateLimiter doesn't expose this information which means we will have to build the machinery to collect it, which I imagine will involve capturing timings inside critical pieces of code. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 3h 10m > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17736213#comment-17736213 ] Stefan Miklosovic commented on CASSANDRA-18305: --- So if I understood that correctly, you would like to see just this and we would leave out 5m and 1m metrics? {code:java} 15 minute rate7.2/minute (that would be 0.12/s * 60s) mean rate 6/minute (that would be 0.01/s * 60s) {code} I am still not completely convinced this makes sense to report but anyway Maybe we should omit them for now and dedicate a separate ticket where we take a look more holistically instead of introducing something which will be later down the road hard to get rid of as we are adding to nodetool output which is very delicate matter. We can always and something if it turns out to be necessary. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 3h 10m > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17736203#comment-17736203 ] Brad Schoening commented on CASSANDRA-18305: [~smiklosovic] I suggested these rate metrics because they are some of the very few JMX compaction metrics available. They're the best approximation of a trend line we have, but I agree, they're not all that useful. The mean rate is since the server started, and I could see comparing the current rate to the mean could be somewhat useful. It would be much better to have the Meter measure compaction throughput in Mbps. Especially since that value has a throttle in the cassandra.yaml. [~brandon.williams] why was it difficult to have compaction throughput (Mbps) as a JMX Meter? Normalizing the values to minutes would be good. It's per/sec only because that's what a Dropwizard Meter measures. Given the skimpy metrics we have, having just a) mean and b)15 minute rate reported per/minute. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 3h 10m > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17736077#comment-17736077 ] Stefan Miklosovic commented on CASSANDRA-18305: --- Thanks [~brandon.williams], I will wait for [~bschoeni] with the answer and we might finally approach the merge. I am in favor of removing these particular metrics but maybe Brad has some reasoning behind that. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 3h 10m > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17736076#comment-17736076 ] Brandon Williams commented on CASSANDRA-18305: -- Whatever we decided to do there, CI looks good so I'm +1. I agree that the number of compactions in a given interval isn't super useful without more info like their sizes, which I think tends to mean what you really want to know is the speed, but we found that is difficult to ascertain. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 3h 10m > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17736065#comment-17736065 ] Stefan Miklosovic commented on CASSANDRA-18305: --- [~bschoeni] why do you think this is important to see? {code} minute rate 0.00/second 5 minute rate0.01/second 15 minute rate 0.12/second mean rate0.01/second {code} or as you put it in the description: {code} pending tasks: 0 compactions completed: 20 1 minute rate:0/second 5 minute rate:2.3/second 15 minute rate: 4.6/second {code} What I mean by that is that I do not see any reason to have these metrics there, what practical information you get from this? If it was a number of how many compactions there were in 15 minutes I might get that, but what is useful in knowing that there was 0.12 compactions per second over 15 minutes? > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 3h 10m > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17736006#comment-17736006 ] Stefan Miklosovic commented on CASSANDRA-18305: --- j8 pre-commit https://app.circleci.com/pipelines/github/instaclustr/cassandra/2510/workflows/a5245e10-2fc8-401a-9d5d-e6d171842b13 > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 3h 10m > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17735842#comment-17735842 ] Stefan Miklosovic commented on CASSANDRA-18305: --- ah that was fast! thanks, yes I ll do full circle as mentioned in step 3) > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 3h > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17735834#comment-17735834 ] Brandon Williams commented on CASSANDRA-18305: -- 6) is not a full circle? I'll go ahead and merge ccm and update the tag. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 3h > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17735833#comment-17735833 ] Stefan Miklosovic commented on CASSANDRA-18305: --- I have tested this branch (1) with this dtest branch (2) which was using this ccm branch (3) and I run dtests against (1) here (4,5). (2) needs to go to dtest as there is one test which is using a custom wait_for_compaction method instead the one on node in ccm. All changes were run against the current trunk (6) so I think we are good for other branches too as nothing has changed but dtest / ccm which was tested in the current trunk already. (1) https://github.com/apache/cassandra/pull/2432/commits (2) https://github.com/apache/cassandra-dtest/pull/222/files (3) https://github.com/riptano/ccm/pull/752 (4) j8_dtest https://app.circleci.com/pipelines/github/instaclustr/cassandra/2507/workflows/9b6867f2-a87e-4fd7-a833-fc66d5482a5b/jobs/59220 (5) j11_dtest https://app.circleci.com/pipelines/github/instaclustr/cassandra/2507/workflows/9b6867f2-a87e-4fd7-a833-fc66d5482a5b/jobs/59221 (6) https://app.circleci.com/pipelines/github/instaclustr/cassandra/2509/workflows/3c3c43cf-0626-4eeb-b76e-30711fcf7996/jobs/59336 > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 3h > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17735437#comment-17735437 ] Stefan Miklosovic commented on CASSANDRA-18305: --- Aha, I remember now, that looks simple. I will do couple builds for various branches to verify it works OK and I will update this ticket. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 2h 40m > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17735434#comment-17735434 ] Brandon Williams commented on CASSANDRA-18305: -- We use the [cassandra-test|https://github.com/apache/cassandra-dtest/blob/trunk/requirements.txt#L12] tag, so we can change ccm easily. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 2h 40m > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17735433#comment-17735433 ] Stefan Miklosovic commented on CASSANDRA-18305: --- If you expect / see some upcoming changes in ccm in a foreseeable future, we may postpone this ticket so it will be done all at once and we do not need to deal with ccm / releases unnecessarily multiple times. I am just thinking about sneaking this in once we are going to do some changes to ccm anyway. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 2h 40m > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17735432#comment-17735432 ] Brandon Williams commented on CASSANDRA-18305: -- Yeah, that looks bad. If we go with fixing the ccm regex, then we should only put this in trunk, but I think that's the best compromise. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 2h 40m > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17735429#comment-17735429 ] Stefan Miklosovic commented on CASSANDRA-18305: --- That would make compactionstats output very ugly and inconsistent. It would be like: {code:java} $ ./bin/nodetool compactionstats concurrent compactors2 pending tasks: 0 ks tb 3 ks tb2 5 compactions completed51 data compacted 8.99 KiB compactions aborted 0 compactions reduced 0 sstables dropped from compaction 0 minute rate 0.60/second 5 minute rate0.60/second 15 minute rate 0.60/second mean rate0.42/second compaction throughput (MBps) 64.0 {code} We may play with the output so it would be on the first line but still ... my ocd is kicking in pretty hard here > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 2h 40m > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17735425#comment-17735425 ] Brandon Williams commented on CASSANDRA-18305: -- Yet another option is restoring the expected behavior and not having to change anything else though, isn't it? I don't think breaking ccm is worth some spaces and a colon. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 2h 40m > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17735418#comment-17735418 ] Stefan Miklosovic commented on CASSANDRA-18305: --- There are two approaches in general 1) we fix ccm, this would be needed (1) The advantage of 1) is that we do not need to change anything in dtests at all and it will be all backward compatible and we do not break anything for anybody and we can change the output in nodetool in trunk. Disadvantage is that we need to do this in ccm, I am not sure how this works in regards of dtests / ccm we are using there, some images need to be released with the patch, no? 2) we do not fix it in ccm, that means we need to fix it in dtests like this (2) The downside of that is that we break ccm because ccm will not be aligned with the work in trunk anymore. The third option is to completely bypass this by not reading the output but reading a metric from JMX. That solution would be on par when it comes to ccm changes / releases so it is basically same as 1) but the advantage is that we are not dependent on some nodetool ouput (!!!) anymore. As I mentioned above, that metric is there from 2.1 beta1. Thanks for looking into it in advance! Really appreciate it. (1) https://github.com/riptano/ccm/pull/752/files (2) https://github.com/smiklosovic/cassandra-dtest/tree/CASSANDRA-18305 > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 2h 40m > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17735412#comment-17735412 ] Brandon Williams commented on CASSANDRA-18305: -- I haven't had a chance to look yet, but how hard is restoring the behavior that ccm is expecting? We aren't the only consumers of ccm so breaking something there could have greater impact, and working around it in the dtests isn't a friendly solution either. bq. BTW do we really want to see this in 4.1 and trunk? I would go with trunk only. If we must break ccm, then definitely trunk only. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 2h 40m > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17735411#comment-17735411 ] Stefan Miklosovic commented on CASSANDRA-18305: --- Another solution is to rewrite logic in ccm to use the metric "PendingTasks" via JMX in CompactionMetrics which returns an integer, that is fairly simple. That metric is there from 2.1 beta1. I humbly think that using nodetool compactionstats command and parsing the output in ccm is very unfortunate approach which backfired here. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 2h 40m > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17734474#comment-17734474 ] Stefan Miklosovic commented on CASSANDRA-18305: --- The build failed because our dtests are using this method (1) which is waiting for an output which never happens as we changed it. It is without semicolon and alignment is different as well. We either fix this directly in ccm (that is rather tricky / too much work with the release etc) or we just fix this in the tests. {code} $ find . -type f -name '*.py' | grep -v './src/' | grep -v './lib/' | grep -v './venv/' | xargs grep wait_for_compactions ./offline_tools_test.py:self.wait_for_compactions(node1) ./offline_tools_test.py:def wait_for_compactions(self, node): ./offline_tools_test.py:self.wait_for_compactions(node1) ./repair_tests/incremental_repair_test.py: cluster.wait_for_compactions() ./repair_tests/incremental_repair_test.py:cluster.wait_for_compactions() ./compaction_test.py:node1.wait_for_compactions() ./compaction_test.py:node1.wait_for_compactions() ./compaction_test.py:node1.wait_for_compactions() ./sstableutil_test.py:node.wait_for_compactions() ./secondary_indexes_test.py:cluster.wait_for_compactions() ./secondary_indexes_test.py:cluster.wait_for_compactions() ./disk_balance_test.py:node.wait_for_compactions() ./disk_balance_test.py:node.wait_for_compactions() {code} The approach I came up with is this (2) Basically, I changed the regexp to optionally include semicolon, then some spaces (does not matter how many) and then "0". This should be backward compatible with all previous versions. I have also changed all places where this check is done by extracting the method to dtest.py branch 4.0 https://github.com/instaclustr/cassandra/tree/CASSANDRA-18305-4.0 j8 pre-commit https://app.circleci.com/pipelines/github/instaclustr/cassandra/2477/workflows/311d0d13-5e8a-42f5-aa64-7f294456ef6b j11 pre-commit https://app.circleci.com/pipelines/github/instaclustr/cassandra/2477/workflows/3b2c582d-d259-4227-88e9-06b67be18d09 I ll provide other builds if this is agreed on. BTW do we really want to see this in 4.1 and trunk? I would go with trunk only. (1) https://github.com/riptano/ccm/blob/master/ccmlib/node.py#L981 (2) https://github.com/smiklosovic/cassandra-dtest/tree/CASSANDRA-18305 > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 2h 40m > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17734041#comment-17734041 ] Stefan Miklosovic commented on CASSANDRA-18305: --- Thanks [~manish.c.ghildi...@gmail.com] for the changes. The last thing I did to 4.0 branch is this (1). Basically, we were using two TableBuilders instead of just one. I consolidated this so all output is rendered into one table. I accommodated tests as well. The patch is not applying cleanly from to 4.1 so I took care of that too. I am going to build 4.0 and 4.1 patches in CircleCI and I will let you know how it goes. I think we have everything necessary to finish this. https://github.com/instaclustr/cassandra/commit/1458de87d994cf5735dc452568556c54e2f6228e > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 2h 40m > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17733767#comment-17733767 ] Manish Ghildiyal commented on CASSANDRA-18305: -- PR updated for 4.0: [PR|https://github.com/apache/cassandra/pull/2375/files] PR created for 4.1:[PR|https://github.com/apache/cassandra/pull/2422/files] > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 2h 40m > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17732547#comment-17732547 ] Brandon Williams commented on CASSANDRA-18305: -- bq. As of now I do not know about any way how to get "the realtime compaction speed" so I think it is better if we omit this completely. I looked into guava's RateLimiter and... I agree. This doesn't work like I thought and it's going to be painful to get this information, it's not readily accessible right now. Thanks for catching this! > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 2.5h > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17732500#comment-17732500 ] Stefan Miklosovic commented on CASSANDRA-18305: --- [~manish.c.ghildi...@gmail.com] great! Thank you for your perseverance. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 2.5h > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17732493#comment-17732493 ] Manish Ghildiyal commented on CASSANDRA-18305: -- [~smiklosovic] , I agree with your point regarding compaction throughput and I have raised it on other PR(for 5.0). I would integrate above recommended changes, and create a separate PR for 4.1 too. Also, would make these changes for 5.0 PR too. I would be able to do it on weekend though. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 2.5h > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17732490#comment-17732490 ] Stefan Miklosovic commented on CASSANDRA-18305: --- After looking more closer to it, I do not think that 'compaction throughput ratio' is computed correctly or that figure even make sense. I think the original idea behind this was what if a node is compacting SSTables, there would be a percentage, like 20%, if it compacts with speed around 12 MBps. However, if you look into CompactionManager, there is: private final RateLimiter compactionRateLimiter = RateLimiter.create(Double.MAX_VALUE); And then we get the rate like: {code} public double getCompactionRate() { return compactionRateLimiter.getRate(); } {code} This should be changed to {code} public double getCompactionRate() { return getRateLimiter().getRate(); } {code} because getRateLimiter() is setting the rates based on how DatabaseDescriptor is configured: {code} public RateLimiter getRateLimiter() { setRate(DatabaseDescriptor.getCompactionThroughputMbPerSec()); return compactionRateLimiter; } {code} and then {code} public void setRate(final double throughPutMbPerSec) { double throughput = throughPutMbPerSec * 1024.0 * 1024.0; // if throughput is set to 0, throttling is disabled if (throughput == 0 || StorageService.instance.isBootstrapMode()) throughput = Double.MAX_VALUE; if (compactionRateLimiter.getRate() != throughput) compactionRateLimiter.setRate(throughput); } {code} Secondly, once this is done, what did we actually obtained by calling CompactionManagerMBean and having `getCompactionRate()` as a double? We just got the maximum throughput allowed. But this does not mean that we got _the actual throughput_. We just got the max possible. So further computing the percentage will not make sense because we always get 100%. This in the patch {code} double configured = probe.getCompactionThroughput(); double actual = probe.getCompactionRate() / (1024 * 1024); {code} these are basically same figures and they will not change over time, regardless how "fast" we compact. As of now I do not know about any way how to get "the realtime compaction speed" so I think it is better if we omit this completely. There are also these metrics which are not covered currently: BytesCompacted CompactionsAborted CompactionsReduced SSTablesDroppedFromCompaction BytesCompacted metric tracks the total amount of bytes Cassandra node compacted since its start. We may probably convert it to something human readable instead of just "bytes" which are hard to parse to something meaningful over time. CompactionsAborted is self-explanatory. CompactionsReduced and SSTablesDroppedFromCompaction are interesting metrics. They are set in CompactionTask.buildCompactionCandidatesForAvailableDiskSpace. That method checks if it has enough disk space for the compaction. CompactionsReduced is increased if it evaluates that some SSTables will be skipped from compaction because if it was compacted it would not fit into the disk. The second metric, SSTablesDroppedFromCompaction, is a counter which adds the number of sstables which were left out from a particular compaction because they would not fit into the disk. This all should be in the output too. CompactionsAborted, CompactionsReduced and SSTablesDroppedFromCompaction where not added in NodeProbe.getCompactionMetric so I added them there too. So, right now, it looks like this: {code} $ ./bin/nodetool compactionstats concurrent compactors2 pending compaction tasks 0 ks tb 3 ks tb2 5 compactions completed51 data compacted 8.99 KiB compactions aborted 0 compactions reduced 0 sstables dropped from compaction 0 minute rate 0.60/second 5 minute rate0.60/second 15 minute rate 0.60/second mean rate0.42/second compaction throughput (MBps) 64.0 {code} I pushed the changes to the same PR (1) as this commit (2) [~manish.c.ghildi...@gmail.com] feel free to take it again and just tweak the tests. I know it might be little bit frustrating we are returning to this but it is how it is ... Please tell us you are ok with this, if you are busy we may help you with that too. (1) https://github.com/apache/cassandra/pull/2393/files (2) https://github.com/apache/cassandra/pull/2393/commits/02b128c1f17f5880ced21007389366385a52ecf8 > Enhance nodetool compactionstats with existing MBean metrics >
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=1773#comment-1773 ] Brandon Williams commented on CASSANDRA-18305: -- Unfortunately 4.0 does not cleanly merge into 4.1 now. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 2.5h > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17731173#comment-17731173 ] Manish Ghildiyal commented on CASSANDRA-18305: -- Tablebuilder changes integrated for 4.0: [PR|https://github.com/apache/cassandra/pull/2375] Formating fixes made for 5.0: [PR|https://github.com/apache/cassandra/pull/2333] > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 1h 40m > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17729918#comment-17729918 ] Manish Ghildiyal commented on CASSANDRA-18305: -- Ok, would take care of comments. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 1h 40m > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17729800#comment-17729800 ] Stefan Miklosovic commented on CASSANDRA-18305: --- Hi [~manish.c.ghildi...@gmail.com] first of all thank your for your contribution. I look at 4.0 patch and what I would prefer to see is the usage of TableBuilder instead of just printing lines to output. TableBuilder is used across the codebase in nodetool and it is just better way to handle the output as it is automatically aligned depending on the width of each line. For example, instead of this (current state of 4.0 branch) {code} 2 concurrent compactors, 0 pending tasks compactions completed: 80 minute rate:.12/second 5 minute rate:.54/second 15 minute rate:.70/second Mean rate:.14/second compaction throughput ratio: 64.0 MBps / 64.0 MBps (100.0%) {code} Now it looks like this (1) {code} concurrent compactors 2 pending tasks count 0 -- here keyspace table will be printed -- here keyspace2 table2 will be printed -- here keyspace3 table3 will be printed -- here keyspace4 table4 will be printed compactions completed 51 minute rate 0.31/second 5 minute rate 0.38/second 15 minute rate 0.39/second mean rate 0.09/second actual compaction throughput (MBps) throttling disabled (0) {code} or {code} concurrent compactors 2 pending tasks count 0 ks tb 3 ks tb 3 compactions completed 57 minute rate 0.46/second 5 minute rate 1.25/second 15 minute rate 1.47/second mean rate 0.10/second actual compaction throughput (MBps) 64.0 actual compaction throughput ratio 64.0 MBps / 64.0 MBps (100.0%) {code} "--" is just placeholder for any pending compactions on some specific keyspace and table if there are any Do you think it is something which we could use instead? You are welcome to build on top of what I did (you may just git cherry-pick that commit). (1) https://github.com/apache/cassandra/pull/2393/files > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 1h 40m > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17729141#comment-17729141 ] Maxwell Guo commented on CASSANDRA-18305: - Hi [~manish.c.ghildi...@gmail.com] I have left some comments for [pr-2333 |https://github.com/apache/cassandra/pull/2333] , and I think some of the comments also works for another pr. Can you take a look ? Thanks. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 40m > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17728916#comment-17728916 ] Manish Ghildiyal commented on CASSANDRA-18305: -- Backported test to 4.0. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 0.5h > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17728778#comment-17728778 ] Manish Ghildiyal commented on CASSANDRA-18305: -- Ok, would do. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 0.5h > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17728770#comment-17728770 ] Brandon Williams commented on CASSANDRA-18305: -- I suspect CompactionStatsTest will Just Work in 4.0, we can backport that there. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 0.5h > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17728768#comment-17728768 ] Manish Ghildiyal commented on CASSANDRA-18305: -- Because in 4.0 , we dont have any test for console output. I would add one. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 0.5h > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17728767#comment-17728767 ] Brandon Williams commented on CASSANDRA-18305: -- Trunk is (or will become) 5.0, so that is covered. We are intending to fix this from 4.0 forward (see the fixversions set on the ticket.) Ultimately we'll need a PR for 4.0, 4.1 and trunk. If the 4.0 patch applies to 4.1 then I will handle that part, but the 4.0 PR does not have the tests. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.x > > Time Spent: 0.5h > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17728760#comment-17728760 ] Manish Ghildiyal commented on CASSANDRA-18305: -- Ah ok. I made original PR against trunk. Never made any PR against 5.0. Do I need to create one for 5.0 too? > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.0 > > Time Spent: 0.5h > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17728758#comment-17728758 ] Brandon Williams commented on CASSANDRA-18305: -- Yes, I see them in the trunk PR only. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.0 > > Time Spent: 0.5h > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17728756#comment-17728756 ] Manish Ghildiyal commented on CASSANDRA-18305: -- Ok. But can you see the test changes. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.0 > > Time Spent: 0.5h > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17728754#comment-17728754 ] Brandon Williams commented on CASSANDRA-18305: -- That definitely does not appear for me when logged in, or in incognito mode. I think it's because it's in 'pending', but in any case I think that is the best place to get the current compaction rate, since the limiter is what directly controls it. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.0 > > Time Spent: 0.5h > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17728753#comment-17728753 ] Manish Ghildiyal commented on CASSANDRA-18305: -- Body of the comment: *After working a bit more through the code, I am not sure if this is right way of getting compaction rate. I think better/correct way is to get hold of all CompactionInfo instances from ActiveCompactions, and use the data from them to calculate the value.* *Please suggest.* > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.0 > > Time Spent: 0.5h > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17728752#comment-17728752 ] Manish Ghildiyal commented on CASSANDRA-18305: -- Here is the screenshot for comment body. !Screenshot 2023-06-02 at 7.33.49 PM.png! > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.0 > > Attachments: Screenshot 2023-06-02 at 7.33.49 PM.png > > Time Spent: 0.5h > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17728751#comment-17728751 ] Brandon Williams commented on CASSANDRA-18305: -- That doesn't take me to a comment, it takes me to the definition of getCompactionRate. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.0 > > Time Spent: 0.5h > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17728749#comment-17728749 ] Manish Ghildiyal commented on CASSANDRA-18305: -- [Link to comment|https://github.com/apache/cassandra/pull/2333/files#diff-f7dd0237c343649f70b7ec9fefd7f6941a40b5164fd6063dce00fc09d2c234a7R2338-R2341] > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.0 > > Time Spent: 0.5h > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17728747#comment-17728747 ] Brandon Williams commented on CASSANDRA-18305: -- Your [commit with the test|https://github.com/apache/cassandra/pull/2333/commits/8f53e9d6a263c55c8fa39cc020f855d9d66021cc] is amended to a [merge commit|https://github.com/apache/cassandra/commit/8f53e9d6a263c55c8fa39cc020f855d9d66021cc] so it's not easy to separate and I can't find your comment, can you repeat it here? > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.0 > > Time Spent: 0.5h > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17728391#comment-17728391 ] Manish Ghildiyal commented on CASSANDRA-18305: -- Yea, I have added one for main branch locally. But I have left a comment for that PR. Waiting for response on that before I push my test. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.0 > > Time Spent: 0.5h > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17728389#comment-17728389 ] Brandon Williams commented on CASSANDRA-18305: -- Looks like we still need the unit tests. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.0 > > Time Spent: 0.5h > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17726780#comment-17726780 ] Manish Ghildiyal commented on CASSANDRA-18305: -- Created [PR|https://github.com/apache/cassandra/pull/2375] for cassandra 4.0 branch. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.0 > > Time Spent: 20m > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17726586#comment-17726586 ] Brandon Williams commented on CASSANDRA-18305: -- I suspect you switched to that branch from another that was previously built, so you should run 'ant realclean' to clean out the old stuff. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.0 > > Time Spent: 10m > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17726577#comment-17726577 ] Manish Ghildiyal commented on CASSANDRA-18305: -- Hi guys, I checked out cassandra-4.1 branch, and I try to build it. But I see many type compatibility issues, like this one: *_error: incompatible types: org.apache.cassandra.io.sstable.format.big.RowIndexEntry cannot be converted to org.apache.cassandra.db.RowIndexEntry_* I am on jdk 8. With no local change, I expect it to build without any problem. Seems I am missing something in local workspace. Any lead here? > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.0 > > Time Spent: 10m > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17726235#comment-17726235 ] Manish Ghildiyal commented on CASSANDRA-18305: -- [~bschoeni] , yes. I would work this weekend. I would make changes for a new PR in 4.1. And also try for unit tests mentioned by [~mmuzaf]. Is it needed immediately? > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.0 > > Time Spent: 10m > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17726229#comment-17726229 ] Brad Schoening commented on CASSANDRA-18305: [~manish.c.ghildi...@gmail.com] have you had a chance to review the above comments? > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.0 > > Time Spent: 10m > Remaining Estimate: 0h > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17724298#comment-17724298 ] Brandon Williams commented on CASSANDRA-18305: -- I don't know why (I think maybe they were amended to a merge commit) but your updates weren't integrating with your PR, so I extracted everything to a single commit [here|https://github.com/driftx/cassandra/commit/4bcbbc189952f1be037601e4c5b532407de5c2a2]. Since this ticket targets 4.0 and 4.1 and this patch doesn't apply cleanly to them, we'll need at least one more PR for 4.0, and another if the same patch doesn't apply to 4.1. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.0 > > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17723155#comment-17723155 ] Brandon Williams commented on CASSANDRA-18305: -- You should set this ticket to Patch Available if you are ready for it to be reviewed so that potential reviewers may find it. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.0 > > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17723154#comment-17723154 ] Manish Ghildiyal commented on CASSANDRA-18305: -- Any update on PR? > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.0 > > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17722456#comment-17722456 ] Manish Ghildiyal commented on CASSANDRA-18305: -- Created [PR|https://github.com/apache/cassandra/pull/2333] > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.0 > > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17722369#comment-17722369 ] Manish Ghildiyal commented on CASSANDRA-18305: -- I see there is CompactionMetrics.completedTasks, as well as CompactionManagerMBean.getCompactionHistory whose count too gives us a hint of number of compacted tasks. But both have different values. I guess CompactionMetrics.completedTasks is a better measure of completed compactions. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.0 > > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17722067#comment-17722067 ] Manish Ghildiyal commented on CASSANDRA-18305: -- I see that a call to NodeProbe:getCompactionThroughputMebibytesAsDouble eventually reads Config.compaction_throughput.toMebibytesPerSecond(), ie what is configured. Is this expected behaviour? I assume this NodeProbe must fetch me current compaction, not what is configured. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.0 > > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17722047#comment-17722047 ] Brad Schoening commented on CASSANDRA-18305: [~manish.c.ghildi...@gmail.com] yes, correct, current compaction rate / compaction_throughput_mb_per_sec. It's also possible that compaction_throughput_mb_per_sec = 0 in which case its unlimited. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.0 > > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17722002#comment-17722002 ] Manish Ghildiyal commented on CASSANDRA-18305: -- *compaction throughput 13.2 MBps / 16 MBps (82.5%)* I guess 16 MBps here is the value mentioned in cassandra.yaml. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Assignee: Manish Ghildiyal >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.0 > > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17720643#comment-17720643 ] Brad Schoening commented on CASSANDRA-18305: [~manish.c.ghildi...@gmail.com] That would be great, please feel free to assign this to yourself if you are able to work on it. I'm not sure what you are looking for in terms of tests. You could likely run cassandra-stress to generate a workload and then look at the MBeans. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.0 > > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17720499#comment-17720499 ] Manish Ghildiyal commented on CASSANDRA-18305: -- I can look into it. Knowing a test that can be debugged to understand what happens where would be of great help. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Priority: Normal > Fix For: 4.0.x, 4.1.x, 5.0 > > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured, perhaps simply add to existing 'pending tasks' line: > {quote}4 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17698228#comment-17698228 ] Stefan Miklosovic commented on CASSANDRA-18305: --- I do not plan to work on that. Feel free to take it. > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Priority: Normal > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured > {quote}concurrent compactions permitted: 2 > {quote} > or perhaps simply add to existing 'pending tasks' line: > {quote}2 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org
[jira] [Commented] (CASSANDRA-18305) Enhance nodetool compactionstats with existing MBean metrics
[ https://issues.apache.org/jira/browse/CASSANDRA-18305?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17698140#comment-17698140 ] maxwellguo commented on CASSANDRA-18305: nobody working on this ? [~smiklosovic] it seems that you have assigned this to you ? > Enhance nodetool compactionstats with existing MBean metrics > > > Key: CASSANDRA-18305 > URL: https://issues.apache.org/jira/browse/CASSANDRA-18305 > Project: Cassandra > Issue Type: Improvement > Components: Tool/nodetool >Reporter: Brad Schoening >Priority: Normal > > Nodetool compactionstats reports only on active compactions, if nothing is > active, you see only: > {quote}$nodetool compactionstats > pending tasks: 0 > {quote} > but in the MBean Compaction/TotalCompactionsCompleted there are recent > statistic in events/second for: > * Count > * FifteenMinueRate > * FiveMinueRate > * MeanRate > * OneMinuteRate > 1) It would be useful to see in addition: > {quote}pending tasks: 0 > compactions completed: 20 > 1 minute rate: 0/second > 5 minute rate: 2.3/second > 15 minute rate: 4.6/second > {quote} > 2) Since compaction_throughput_mb_per_sec is a throttling parameter in > cassandra.yaml (default 64 MBps), it would be nice to show the actual > compaction throughput and be able to observe if you're close to the limit. > I.e., > {quote}compaction throughput 13.2 MBps / 16 MBps (82.5%) > {quote} > 3) for completness, compactionstats should list the number of concurrent > compactors configured > {quote}concurrent compactions permitted: 2 > {quote} > or perhaps simply add to existing 'pending tasks' line: > {quote}2 concurrent compactors, 0 pending tasks > {quote} -- This message was sent by Atlassian Jira (v8.20.10#820010) - To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org