[jira] [Commented] (YARN-3870) Providing raw container request information for fine scheduling
[ https://issues.apache.org/jira/browse/YARN-3870?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15076642#comment-15076642 ] Arun Suresh commented on YARN-3870: --- +1 to having the AM create the ID (which I guess is required if the AM is to co-relate a resourcereq with a container response). Can we set the ID that is something of a combination of {{app_attempt_id + seq_id}}, where seq_id is incremented monotonically per app attempt ? to distinguish between different app attempts. Also, I guess the outstanding requests for an app needs to be stored in the state store so subsequent app attempts can be intimated of unfullfulled reqs > Providing raw container request information for fine scheduling > --- > > Key: YARN-3870 > URL: https://issues.apache.org/jira/browse/YARN-3870 > Project: Hadoop YARN > Issue Type: Sub-task > Components: api, applications, capacityscheduler, fairscheduler, > resourcemanager, scheduler, yarn >Reporter: Lei Guo >Assignee: Karthik Kambatla > > Currently, when AM sends container requests to RM and scheduler, it expands > individual container requests into host/rack/any format. For instance, if I > am asking for container request with preference "host1, host2, host3", > assuming all are in the same rack rack1, instead of sending one raw container > request to RM/Scheduler with raw preference list, it basically expand it to > become 5 different objects with host1, host2, host3, rack1 and any in there. > When scheduler receives information, it basically already lost the raw > request. This is ok for single container request, but it will cause trouble > when dealing with multiple container requests from the same application. > Consider this case: > 6 hosts, two racks: > rack1 (host1, host2, host3) rack2 (host4, host5, host6) > When application requests two containers with different data locality > preference: > c1: host1, host2, host4 > c2: host2, host3, host5 > This will end up with following container request list when client sending > request to RM/Scheduler: > host1: 1 instance > host2: 2 instances > host3: 1 instance > host4: 1 instance > host5: 1 instance > rack1: 2 instances > rack2: 2 instances > any: 2 instances > Fundamentally, it is hard for scheduler to make a right judgement without > knowing the raw container request. The situation will get worse when dealing > with affinity and anti-affinity or even gang scheduling etc. > We need some way to provide raw container request information for fine > scheduling purpose. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Assigned] (YARN-3870) Providing raw container request information for fine scheduling
[ https://issues.apache.org/jira/browse/YARN-3870?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Karthik Kambatla reassigned YARN-3870: -- Assignee: Karthik Kambatla > Providing raw container request information for fine scheduling > --- > > Key: YARN-3870 > URL: https://issues.apache.org/jira/browse/YARN-3870 > Project: Hadoop YARN > Issue Type: Sub-task > Components: api, applications, capacityscheduler, fairscheduler, > resourcemanager, scheduler, yarn >Reporter: Lei Guo >Assignee: Karthik Kambatla > > Currently, when AM sends container requests to RM and scheduler, it expands > individual container requests into host/rack/any format. For instance, if I > am asking for container request with preference "host1, host2, host3", > assuming all are in the same rack rack1, instead of sending one raw container > request to RM/Scheduler with raw preference list, it basically expand it to > become 5 different objects with host1, host2, host3, rack1 and any in there. > When scheduler receives information, it basically already lost the raw > request. This is ok for single container request, but it will cause trouble > when dealing with multiple container requests from the same application. > Consider this case: > 6 hosts, two racks: > rack1 (host1, host2, host3) rack2 (host4, host5, host6) > When application requests two containers with different data locality > preference: > c1: host1, host2, host4 > c2: host2, host3, host5 > This will end up with following container request list when client sending > request to RM/Scheduler: > host1: 1 instance > host2: 2 instances > host3: 1 instance > host4: 1 instance > host5: 1 instance > rack1: 2 instances > rack2: 2 instances > any: 2 instances > Fundamentally, it is hard for scheduler to make a right judgement without > knowing the raw container request. The situation will get worse when dealing > with affinity and anti-affinity or even gang scheduling etc. > We need some way to provide raw container request information for fine > scheduling purpose. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (YARN-3870) Providing raw container request information for fine scheduling
[ https://issues.apache.org/jira/browse/YARN-3870?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15076629#comment-15076629 ] Karthik Kambatla commented on YARN-3870: bq. Karthik Kambatla, I am thinking whether we also need some update for the response part to correlate it with the ResourceRequest ID. As the scheduling is asynchronous, AM will also need to know the relation between response and request. bq. If the AM doesn't add an ID, the RM could add one. Or, we could have the RM add the IDs and return them to the AM for help with book keeping. Thought more about this. Since one AllocateRequest could have multiple ResourceRequests, the protocol becomes quite complicated if the RM creates an ID instead of the AM. How about we expect the AM to set this ID? If the AM doesn't set, we treat the requests the same way we do today (ID = -1). In the AllocateResponse, the RM could send the last received ResourceRequest. The AM could look at this ACK to see if it has to resend the requests? The AMRMClient and MR-AM could be updated to do this. > Providing raw container request information for fine scheduling > --- > > Key: YARN-3870 > URL: https://issues.apache.org/jira/browse/YARN-3870 > Project: Hadoop YARN > Issue Type: Sub-task > Components: api, applications, capacityscheduler, fairscheduler, > resourcemanager, scheduler, yarn >Reporter: Lei Guo > > Currently, when AM sends container requests to RM and scheduler, it expands > individual container requests into host/rack/any format. For instance, if I > am asking for container request with preference "host1, host2, host3", > assuming all are in the same rack rack1, instead of sending one raw container > request to RM/Scheduler with raw preference list, it basically expand it to > become 5 different objects with host1, host2, host3, rack1 and any in there. > When scheduler receives information, it basically already lost the raw > request. This is ok for single container request, but it will cause trouble > when dealing with multiple container requests from the same application. > Consider this case: > 6 hosts, two racks: > rack1 (host1, host2, host3) rack2 (host4, host5, host6) > When application requests two containers with different data locality > preference: > c1: host1, host2, host4 > c2: host2, host3, host5 > This will end up with following container request list when client sending > request to RM/Scheduler: > host1: 1 instance > host2: 2 instances > host3: 1 instance > host4: 1 instance > host5: 1 instance > rack1: 2 instances > rack2: 2 instances > any: 2 instances > Fundamentally, it is hard for scheduler to make a right judgement without > knowing the raw container request. The situation will get worse when dealing > with affinity and anti-affinity or even gang scheduling etc. > We need some way to provide raw container request information for fine > scheduling purpose. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (YARN-397) RM Scheduler api enhancements
[ https://issues.apache.org/jira/browse/YARN-397?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15076741#comment-15076741 ] Karthik Kambatla commented on YARN-397: --- I believe the intent of this umbrella JIRA was to keep track of all scheduler API changes before Yarn went beta. Would it make more sense to go through individual JIRAs listed here, convert them to issues and close this umbrella JIRA? [~acmurthy], [~vinodkv]? > RM Scheduler api enhancements > - > > Key: YARN-397 > URL: https://issues.apache.org/jira/browse/YARN-397 > Project: Hadoop YARN > Issue Type: Improvement >Reporter: Arun C Murthy > > Umbrella jira tracking enhancements to RM apis. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Updated] (YARN-2885) Create AMRMProxy request interceptor for distributed scheduling decisions for queueable containers
[ https://issues.apache.org/jira/browse/YARN-2885?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ] Arun Suresh updated YARN-2885: -- Attachment: YARN-2885-yarn-2877.v5.patch Updating patch with javadocs and addressing some of [~leftnoteasy]'s suggestions > Create AMRMProxy request interceptor for distributed scheduling decisions for > queueable containers > -- > > Key: YARN-2885 > URL: https://issues.apache.org/jira/browse/YARN-2885 > Project: Hadoop YARN > Issue Type: Sub-task > Components: nodemanager, resourcemanager >Reporter: Konstantinos Karanasos >Assignee: Arun Suresh > Attachments: YARN-2885-yarn-2877.001.patch, > YARN-2885-yarn-2877.002.patch, YARN-2885-yarn-2877.full-2.patch, > YARN-2885-yarn-2877.full-3.patch, YARN-2885-yarn-2877.full.patch, > YARN-2885-yarn-2877.v4.patch, YARN-2885-yarn-2877.v5.patch, > YARN-2885_api_changes.patch > > > We propose to add a Local ResourceManager (LocalRM) to the NM in order to > support distributed scheduling decisions. > Architecturally we leverage the RMProxy, introduced in YARN-2884. > The LocalRM makes distributed decisions for queuable containers requests. > Guaranteed-start requests are still handled by the central RM. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (YARN-3870) Providing raw container request information for fine scheduling
[ https://issues.apache.org/jira/browse/YARN-3870?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15076743#comment-15076743 ] Karthik Kambatla commented on YARN-3870: Was fleshing this out further. The number of IDs and hence ResourceRequests could be O(num. outstanding containers) which could pose problems as outlined in YARN-371. In fact, this JIRA is a duplicate of YARN-371: may be, we should close this and continue the discussion there. > Providing raw container request information for fine scheduling > --- > > Key: YARN-3870 > URL: https://issues.apache.org/jira/browse/YARN-3870 > Project: Hadoop YARN > Issue Type: Sub-task > Components: api, applications, capacityscheduler, fairscheduler, > resourcemanager, scheduler, yarn >Reporter: Lei Guo >Assignee: Karthik Kambatla > > Currently, when AM sends container requests to RM and scheduler, it expands > individual container requests into host/rack/any format. For instance, if I > am asking for container request with preference "host1, host2, host3", > assuming all are in the same rack rack1, instead of sending one raw container > request to RM/Scheduler with raw preference list, it basically expand it to > become 5 different objects with host1, host2, host3, rack1 and any in there. > When scheduler receives information, it basically already lost the raw > request. This is ok for single container request, but it will cause trouble > when dealing with multiple container requests from the same application. > Consider this case: > 6 hosts, two racks: > rack1 (host1, host2, host3) rack2 (host4, host5, host6) > When application requests two containers with different data locality > preference: > c1: host1, host2, host4 > c2: host2, host3, host5 > This will end up with following container request list when client sending > request to RM/Scheduler: > host1: 1 instance > host2: 2 instances > host3: 1 instance > host4: 1 instance > host5: 1 instance > rack1: 2 instances > rack2: 2 instances > any: 2 instances > Fundamentally, it is hard for scheduler to make a right judgement without > knowing the raw container request. The situation will get worse when dealing > with affinity and anti-affinity or even gang scheduling etc. > We need some way to provide raw container request information for fine > scheduling purpose. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (YARN-3870) Providing raw container request information for fine scheduling
[ https://issues.apache.org/jira/browse/YARN-3870?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15076746#comment-15076746 ] Arun Suresh commented on YARN-3870: --- Hmmm... not sure entirely sure the argument holds, or I might be missing something. For eg. even without an ID, what happens today if an application makes each request with a different Priority ? > Providing raw container request information for fine scheduling > --- > > Key: YARN-3870 > URL: https://issues.apache.org/jira/browse/YARN-3870 > Project: Hadoop YARN > Issue Type: Sub-task > Components: api, applications, capacityscheduler, fairscheduler, > resourcemanager, scheduler, yarn >Reporter: Lei Guo >Assignee: Karthik Kambatla > > Currently, when AM sends container requests to RM and scheduler, it expands > individual container requests into host/rack/any format. For instance, if I > am asking for container request with preference "host1, host2, host3", > assuming all are in the same rack rack1, instead of sending one raw container > request to RM/Scheduler with raw preference list, it basically expand it to > become 5 different objects with host1, host2, host3, rack1 and any in there. > When scheduler receives information, it basically already lost the raw > request. This is ok for single container request, but it will cause trouble > when dealing with multiple container requests from the same application. > Consider this case: > 6 hosts, two racks: > rack1 (host1, host2, host3) rack2 (host4, host5, host6) > When application requests two containers with different data locality > preference: > c1: host1, host2, host4 > c2: host2, host3, host5 > This will end up with following container request list when client sending > request to RM/Scheduler: > host1: 1 instance > host2: 2 instances > host3: 1 instance > host4: 1 instance > host5: 1 instance > rack1: 2 instances > rack2: 2 instances > any: 2 instances > Fundamentally, it is hard for scheduler to make a right judgement without > knowing the raw container request. The situation will get worse when dealing > with affinity and anti-affinity or even gang scheduling etc. > We need some way to provide raw container request information for fine > scheduling purpose. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (YARN-1011) [Umbrella] Schedule containers based on utilization of currently allocated containers
[ https://issues.apache.org/jira/browse/YARN-1011?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15076750#comment-15076750 ] Karthik Kambatla commented on YARN-1011: Forgot to respond to one comment: bq. when terminating opportunistic containers will the RM ask the AM about which containers to kill? Don't think we should. NM --> RM --> AM --> NM is a long communication thread. Our preemption should kick in much faster that that. What do you think of preempting the last opportunistic container that was started, since it is likely that far away from promotion. > [Umbrella] Schedule containers based on utilization of currently allocated > containers > - > > Key: YARN-1011 > URL: https://issues.apache.org/jira/browse/YARN-1011 > Project: Hadoop YARN > Issue Type: New Feature >Reporter: Arun C Murthy > Attachments: yarn-1011-design-v0.pdf > > > Currently RM allocates containers and assumes resources allocated are > utilized. > RM can, and should, get to a point where it measures utilization of allocated > containers and, if appropriate, allocate more (speculative?) containers. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (YARN-1011) [Umbrella] Schedule containers based on utilization of currently allocated containers
[ https://issues.apache.org/jira/browse/YARN-1011?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15076749#comment-15076749 ] Karthik Kambatla commented on YARN-1011: Thanks for chiming in, [~bikassaha]. bq. It is essential to run opportunistic tasks at lower OS cpu priority so that they never obstruct progress of normal tasks. bq. In fact, this is the litmus test for opportunistic scheduling. Good point. Guaranteed containers should get priority for resources: Opportunistic containers should only use left-over resources. We should do this for CPU, disk and network. I am not aware of the latest on disk and network isolation, but we should create sub-tasks for those too. /cc [~vvasudev] bq. Handling opportunistic tasks raises questions on the involvement of the AMs. bq. In that sense it would be instructive to consider opportunistic scheduling in a similar light as preemption. I wasn't sure the AM needs to know a container's execution type: As you mention, this is very similar to preemption. From an AM's standpoint, the container would be preempted if those resources are not available to that application any more. In case of preemption, this can happen if other high priority queues have outstanding demand or the cluster lost a couple of nodes. Here, it is possible Guaranteed containers actually need the resources. In that sense, the AM doesn't have to do anything different for Guaranteed vs Opportunistic containers. Predictability: Allowing applications to specify only Guaranteed containers vs Guaranteed or Opportunistic containers should take care of this. However, between getting no resources and getting opportunistic resources, are there cases where the applications prefer the latter? The applications "should" get guaranteed containers at the same point in time irrespective of whether they use opportunistic resources in the interim. Note that allowing applications to specify whether they are okay with getting opportunistic containers complicates the scheduling - the scheduler needs to look through the higher priority apps that don't allow opportunistic containers before getting to those that need. And, when resources are available on that node, the RM will need to schedule containers for higher priority apps prolonging the duration for which opportunistic containers stay opportunistic. Given this complication, I would prefer we do not involve AMs in the decision-making process. Based on the need and usecases, we could revisit this at a later time. Note that YARN-4335 adds this to ResourceRequest for distributed scheduling, and even there they are not entirely sure if it needs to be a part. bq. does the AM need to know that a newly allocated container was opportunistic. E.g. so that it does not schedule the highest priority work on that container. Valid concern. May be, we should intimate the AM of whether a container is opportunistic, and later when it gets promoted to guaranteed. That said, I am not sure if this is essential to oversubscription being useful. Thoughts on punting it to Phase-2? bq. will opportunistic containers be given only when for containers that are beyond queue capacity such that we dont break any guarantees on their liveliness. ie. an AM will not expect to lose any container that is within its queue capacity but opportunistic containers can be killed at any time. Yes. This probably needs to be clear in the doc. Will update it. bq. will conversion of opportunistic containers to regular containers be automatically done by the RM? By some combination of RM/NM, definitely yes. Initially, I thought the RM can be the only one doing this. The RM could keep track of opportunistic containers in SchedulerNode. Today, we already track launchedContainers. The scheduler could go through this list and promote containers before allocating new containers. Does this add an unnecessary delay in the promotion though? If the scheduler allocated opportunistic containers based on the same prioritization it uses for guaranteed containers, can the NM just promote the oldest opportunistic container running on that node and update the RM accordingly? Another thing to consider here: the promotion process here should work with that in YARN-2877. [~subru], [~kkaranasos], [~asuresh] - is it okay for the NM to automatically promote some opportunistic containers. May be, we could add a flag to the launch context to differentiate between those opportunistic containers that can be automatically promoted vs those that can not be. > [Umbrella] Schedule containers based on utilization of currently allocated > containers > - > > Key: YARN-1011 > URL: https://issues.apache.org/jira/browse/YARN-1011 > Project: Hadoop YARN > Issue Type: New
[jira] [Created] (YARN-4531) org.apache.hadoop.yarn.event.AsyncDispatcher: AsyncDispatcher thread interrupted java.lang.InterruptedException
SHYAM RAMATH created YARN-4531: -- Summary: org.apache.hadoop.yarn.event.AsyncDispatcher: AsyncDispatcher thread interrupted java.lang.InterruptedException Key: YARN-4531 URL: https://issues.apache.org/jira/browse/YARN-4531 Project: Hadoop YARN Issue Type: Bug Components: api Affects Versions: 2.7.1 Environment: Windows environment Reporter: SHYAM RAMATH Priority: Trivial The error occured while executing the Hadoop Job : org.apache.hadoop.yarn.event.AsyncDispatcher: AsyncDispatcher thread interrupted java.lang.InterruptedException - -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (YARN-3870) Providing raw container request information for fine scheduling
[ https://issues.apache.org/jira/browse/YARN-3870?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15079334#comment-15079334 ] Karthik Kambatla commented on YARN-3870: [~leftnoteasy] - agree that YARN-4485 doesn't necessarily need ID-ing all requests corresponding to one task. The JIRAs are related only due to the data-structures in AppSchedulingInfo: # If we add IDs as discussed here, we don't need any other data-structure changes except adding a timestamp to each ResourceRequest. # If we don't add IDs, we might need to store the resource-requests as MapProviding raw container request information for fine scheduling > --- > > Key: YARN-3870 > URL: https://issues.apache.org/jira/browse/YARN-3870 > Project: Hadoop YARN > Issue Type: Sub-task > Components: api, applications, capacityscheduler, fairscheduler, > resourcemanager, scheduler, yarn >Reporter: Lei Guo >Assignee: Karthik Kambatla > > Currently, when AM sends container requests to RM and scheduler, it expands > individual container requests into host/rack/any format. For instance, if I > am asking for container request with preference "host1, host2, host3", > assuming all are in the same rack rack1, instead of sending one raw container > request to RM/Scheduler with raw preference list, it basically expand it to > become 5 different objects with host1, host2, host3, rack1 and any in there. > When scheduler receives information, it basically already lost the raw > request. This is ok for single container request, but it will cause trouble > when dealing with multiple container requests from the same application. > Consider this case: > 6 hosts, two racks: > rack1 (host1, host2, host3) rack2 (host4, host5, host6) > When application requests two containers with different data locality > preference: > c1: host1, host2, host4 > c2: host2, host3, host5 > This will end up with following container request list when client sending > request to RM/Scheduler: > host1: 1 instance > host2: 2 instances > host3: 1 instance > host4: 1 instance > host5: 1 instance > rack1: 2 instances > rack2: 2 instances > any: 2 instances > Fundamentally, it is hard for scheduler to make a right judgement without > knowing the raw container request. The situation will get worse when dealing > with affinity and anti-affinity or even gang scheduling etc. > We need some way to provide raw container request information for fine > scheduling purpose. -- This message was sent by Atlassian JIRA (v6.3.4#6332)
[jira] [Commented] (YARN-3870) Providing raw container request information for fine scheduling
[ https://issues.apache.org/jira/browse/YARN-3870?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15076775#comment-15076775 ] Wangda Tan commented on YARN-3870: -- [~kasha], bq. Was fleshing this out further. The number of IDs and hence ResourceRequests could be O(num. outstanding containers) which could pose problems as outlined in YARN-371. In fact, this JIRA is a duplicate of YARN-371: may be, we should close this and continue the discussion there. As I mentioned above, I think we shouldn't combine YARN-371 and YARN-4485 together: YARN-4485 is more like an internal change of scheduler to me: Let's say an AM originally requests 1000 container (T1), then AM requests 1200 containers (T2), then after scheduler allocated 100 containers, AM requests 1200 containers again (T3). For the original request, scheduler records: T1, 1000. After T2, scheduler records: T1, 1000; T2, 200. After T3, scheduler records: T1, 900 (scheduler allocates 100 containers); T2, 200; T3, 100. Instead recording timestamps for all resource requests, AM only needs to record timestamp to #pending-requests. And scheduler will "dequeue" from the timestamp to #pending-requests (sorted by time) when container allocated. Like what you said, it will be hard to ask AM to set the ID, but scheduler should easily set it. But this solution needs more work if we want to save these timestamps when RM restart. > Providing raw container request information for fine scheduling > --- > > Key: YARN-3870 > URL: https://issues.apache.org/jira/browse/YARN-3870 > Project: Hadoop YARN > Issue Type: Sub-task > Components: api, applications, capacityscheduler, fairscheduler, > resourcemanager, scheduler, yarn >Reporter: Lei Guo >Assignee: Karthik Kambatla > > Currently, when AM sends container requests to RM and scheduler, it expands > individual container requests into host/rack/any format. For instance, if I > am asking for container request with preference "host1, host2, host3", > assuming all are in the same rack rack1, instead of sending one raw container > request to RM/Scheduler with raw preference list, it basically expand it to > become 5 different objects with host1, host2, host3, rack1 and any in there. > When scheduler receives information, it basically already lost the raw > request. This is ok for single container request, but it will cause trouble > when dealing with multiple container requests from the same application. > Consider this case: > 6 hosts, two racks: > rack1 (host1, host2, host3) rack2 (host4, host5, host6) > When application requests two containers with different data locality > preference: > c1: host1, host2, host4 > c2: host2, host3, host5 > This will end up with following container request list when client sending > request to RM/Scheduler: > host1: 1 instance > host2: 2 instances > host3: 1 instance > host4: 1 instance > host5: 1 instance > rack1: 2 instances > rack2: 2 instances > any: 2 instances > Fundamentally, it is hard for scheduler to make a right judgement without > knowing the raw container request. The situation will get worse when dealing > with affinity and anti-affinity or even gang scheduling etc. > We need some way to provide raw container request information for fine > scheduling purpose. -- This message was sent by Atlassian JIRA (v6.3.4#6332)