[jira] [Commented] (SPARK-26688) Provide configuration of initially blacklisted YARN nodes
[ https://issues.apache.org/jira/browse/SPARK-26688?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16763922#comment-16763922 ] Imran Rashid commented on SPARK-26688: -- I haven't heard any more objections to this, so I think its worth moving this forward. I'll take a closer look at Attila's change and merge next week unless we discover any issues. > Provide configuration of initially blacklisted YARN nodes > - > > Key: SPARK-26688 > URL: https://issues.apache.org/jira/browse/SPARK-26688 > Project: Spark > Issue Type: Improvement > Components: YARN >Affects Versions: 3.0.0 >Reporter: Attila Zsolt Piros >Priority: Major > > Introducing new config for initially blacklisted YARN nodes. > This came up in the apache spark user mailing list: > [http://apache-spark-user-list.1001560.n3.nabble.com/Spark-on-Yarn-is-it-possible-to-manually-blacklist-nodes-before-running-spark-job-td34395.html] > -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org For additional commands, e-mail: issues-h...@spark.apache.org
[jira] [Commented] (SPARK-26688) Provide configuration of initially blacklisted YARN nodes
[ https://issues.apache.org/jira/browse/SPARK-26688?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16753005#comment-16753005 ] Sergey commented on SPARK-26688: Hi Imran, thanks for you reply. "Meanwhile devs start to apply this willy-nilly, as these configs tend to just keep getting built up over time " SLA could mitigate this problem. Every blacklisted node for a specific job slows it down in a long run. Anyway, dev would have to report / communicate with ops to resolve node issue. "Ideally, blacklisting and speculation should be able to prevent that problem" We are going to try out speculation but we are not there yet. > Provide configuration of initially blacklisted YARN nodes > - > > Key: SPARK-26688 > URL: https://issues.apache.org/jira/browse/SPARK-26688 > Project: Spark > Issue Type: Improvement > Components: YARN >Affects Versions: 3.0.0 >Reporter: Attila Zsolt Piros >Priority: Major > > Introducing new config for initially blacklisted YARN nodes. > This came up in the apache spark user mailing list: > [http://apache-spark-user-list.1001560.n3.nabble.com/Spark-on-Yarn-is-it-possible-to-manually-blacklist-nodes-before-running-spark-job-td34395.html] > -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org For additional commands, e-mail: issues-h...@spark.apache.org
[jira] [Commented] (SPARK-26688) Provide configuration of initially blacklisted YARN nodes
[ https://issues.apache.org/jira/browse/SPARK-26688?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16751830#comment-16751830 ] Imran Rashid commented on SPARK-26688: -- OK that is a reasonable request ... but to play devil's advocate, there is a downside to having devs try to use this workaround instead of going through Ops. They never end up sending the feedback to Ops, Ops doesn't ever get metrics on the bad nodes, because nothing even tries to run there. Meanwhile devs start to apply this willy-nilly, as these configs tend to just keep getting built up over time (I've seen cases where configs included 50 GB containers, just because once, 3 years ago, somebody's app needed that much memory, and that config just keeps getting copied forward to everyone). Eventually you've got a hodge-podge of blacklisting in place, and nobody really knows which nodes are actually good or bad. Ideally, blacklisting and speculation should be able to prevent that problem from being so noticeable in the first place, but I can see how reality might be far from that ideal. Nonetheless, on the whole I think I lean slightly in favor of adding this, but would like to hear more opinions. > Provide configuration of initially blacklisted YARN nodes > - > > Key: SPARK-26688 > URL: https://issues.apache.org/jira/browse/SPARK-26688 > Project: Spark > Issue Type: Improvement > Components: YARN >Affects Versions: 3.0.0 >Reporter: Attila Zsolt Piros >Priority: Major > > Introducing new config for initially blacklisted YARN nodes. > This came up in the apache spark user mailing list: > [http://apache-spark-user-list.1001560.n3.nabble.com/Spark-on-Yarn-is-it-possible-to-manually-blacklist-nodes-before-running-spark-job-td34395.html] > -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org For additional commands, e-mail: issues-h...@spark.apache.org
[jira] [Commented] (SPARK-26688) Provide configuration of initially blacklisted YARN nodes
[ https://issues.apache.org/jira/browse/SPARK-26688?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16750498#comment-16750498 ] Sergey commented on SPARK-26688: Hi There! I'm very glad that the community paid attention to my question. Let me try to explain usecase There is 1K nodes cluster and jobs have performance degradation because of a single node. It's rather hard to convince Cluster Ops to decommission node because of "performance degradation". Imagine 10 dev teams chase single ops team for valid reason (node has problems) or because code has a bug or data is skewed or spots on the sun. We can't just decommission node because random dev complains. Simple solution: * rerun failed / delayed job and blacklist "problematic" node in advance. * Report about the problem if job works w/o anomalies. * ops collect complains about node and start to decommission it when "complains threshold" is reached. It's a rather low probability that many loosely coupled teams with loosely coupled jobs complain about a single node. Results * Ops are not spammed with a random requests from devs * devs are not blocked because of the really bad node. * it's very cheap for everyone to "blacklist" node during job submission w/o doing anything to node. > Provide configuration of initially blacklisted YARN nodes > - > > Key: SPARK-26688 > URL: https://issues.apache.org/jira/browse/SPARK-26688 > Project: Spark > Issue Type: Improvement > Components: YARN >Affects Versions: 3.0.0 >Reporter: Attila Zsolt Piros >Priority: Major > > Introducing new config for initially blacklisted YARN nodes. > This came up in the apache spark user mailing list: > [http://apache-spark-user-list.1001560.n3.nabble.com/Spark-on-Yarn-is-it-possible-to-manually-blacklist-nodes-before-running-spark-job-td34395.html] > -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org For additional commands, e-mail: issues-h...@spark.apache.org
[jira] [Commented] (SPARK-26688) Provide configuration of initially blacklisted YARN nodes
[ https://issues.apache.org/jira/browse/SPARK-26688?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16750477#comment-16750477 ] Mridul Muralidharan commented on SPARK-26688: - If this is a legitimate usecase, we should get yarn team to enhance node label support. > Provide configuration of initially blacklisted YARN nodes > - > > Key: SPARK-26688 > URL: https://issues.apache.org/jira/browse/SPARK-26688 > Project: Spark > Issue Type: Improvement > Components: YARN >Affects Versions: 3.0.0 >Reporter: Attila Zsolt Piros >Priority: Major > > Introducing new config for initially blacklisted YARN nodes. > This came up in the apache spark user mailing list: > [http://apache-spark-user-list.1001560.n3.nabble.com/Spark-on-Yarn-is-it-possible-to-manually-blacklist-nodes-before-running-spark-job-td34395.html] > -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org For additional commands, e-mail: issues-h...@spark.apache.org
[jira] [Commented] (SPARK-26688) Provide configuration of initially blacklisted YARN nodes
[ https://issues.apache.org/jira/browse/SPARK-26688?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16748922#comment-16748922 ] Attila Zsolt Piros commented on SPARK-26688: As I know a node only could have one label. So if node label is already used for something else (to group somehow your nodes) this could be a lightweight way to still exclude some YARN nodes. > Provide configuration of initially blacklisted YARN nodes > - > > Key: SPARK-26688 > URL: https://issues.apache.org/jira/browse/SPARK-26688 > Project: Spark > Issue Type: Improvement > Components: YARN >Affects Versions: 3.0.0 >Reporter: Attila Zsolt Piros >Priority: Major > > Introducing new config for initially blacklisted YARN nodes. > This came up in the apache spark user mailing list: > [http://apache-spark-user-list.1001560.n3.nabble.com/Spark-on-Yarn-is-it-possible-to-manually-blacklist-nodes-before-running-spark-job-td34395.html] > -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org For additional commands, e-mail: issues-h...@spark.apache.org
[jira] [Commented] (SPARK-26688) Provide configuration of initially blacklisted YARN nodes
[ https://issues.apache.org/jira/browse/SPARK-26688?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16748915#comment-16748915 ] Mridul Muralidharan commented on SPARK-26688: - What is the usecase for this ? As others have mentioned in the news group, nodelabels is the typical way to prevent (or require) allocation requests to satisfy some constraint. > Provide configuration of initially blacklisted YARN nodes > - > > Key: SPARK-26688 > URL: https://issues.apache.org/jira/browse/SPARK-26688 > Project: Spark > Issue Type: Improvement > Components: YARN >Affects Versions: 3.0.0 >Reporter: Attila Zsolt Piros >Priority: Major > > Introducing new config for initially blacklisted YARN nodes. > This came up in the apache spark user mailing list: > [http://apache-spark-user-list.1001560.n3.nabble.com/Spark-on-Yarn-is-it-possible-to-manually-blacklist-nodes-before-running-spark-job-td34395.html] > -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org For additional commands, e-mail: issues-h...@spark.apache.org
[jira] [Commented] (SPARK-26688) Provide configuration of initially blacklisted YARN nodes
[ https://issues.apache.org/jira/browse/SPARK-26688?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=16748632#comment-16748632 ] Attila Zsolt Piros commented on SPARK-26688: I am working on this. > Provide configuration of initially blacklisted YARN nodes > - > > Key: SPARK-26688 > URL: https://issues.apache.org/jira/browse/SPARK-26688 > Project: Spark > Issue Type: Improvement > Components: YARN >Affects Versions: 3.0.0 >Reporter: Attila Zsolt Piros >Priority: Major > > Introducing new config for initially blacklisted YARN nodes. -- This message was sent by Atlassian JIRA (v7.6.3#76005) - To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org For additional commands, e-mail: issues-h...@spark.apache.org
