Hi Sebastian, How do you change memory settings for Infinispan started via service catalog? The memory settings seem defined in [1], but this is not one of the parameters supported. I guess we want this as parameter? Cheers, [1] https://github.com/infinispan/infinispan-openshift-templates/blob/master/... -- Galder Zamarreño Infinispan, Red Hat

On 22 September 2017 at 13:49, Sebastian Laskawiec <slaskawi(a)redhat.com&gt; wrote: > It's very tricky... > > Memory is adjusted automatically to the container size [1] (of course you > may override it by supplying Xmx or "-n" as parameters [2]). The safe limit > is roughly Xmx=Xms=50% of container capacity (unless you do the off-heap, > that you can squeeze Infinispan much, much more). > > Then there are Limits, Requests and QoS in Kubernetes [3][4]. We are in > bustable memory category so if there is additional memory in the node, we'll > get it. But if not, we won't go below 512 MB (and 500 mCPU). I hope that's a temporary choice of the work in process? Doesn't sound acceptable to address real world requirements.. Infinispan expects users to estimate how much memory they will need - which is hard enough - and then we should at least be able to start a cluster to address the specified need. Being able to rely on 512MB only per node would require lots of nodes even for small data sets, leading to extreme resource waste as each node would consume some non negligible portion of memory just to run the thing.

Thanks, Sanne > > Thanks, > Sebastian > > [1] > https://github.com/jboss-dockerfiles/infinispan/tree/master/server#adjust... > [2] > https://github.com/jboss-dockerfiles/infinispan/blob/master/server/docker... > [3] https://www.youtube.com/watch?v=nWGkvrIPqJ4 > [4] > https://docs.openshift.com/enterprise/3.2/dev_guide/compute_resources.html > > On Fri, Sep 22, 2017 at 2:33 PM Galder Zamarreño <galder(a)redhat.com&gt; wrote: >> >> Hi Sebastian, >> >> How do you change memory settings for Infinispan started via service >> catalog? >> >> The memory settings seem defined in [1], but this is not one of the >> parameters supported. >> >> I guess we want this as parameter? >> >> Cheers, >> >> [1] >> https://github.com/infinispan/infinispan-openshift-templates/blob/master/... >> -- >> Galder Zamarreño >> Infinispan, Red Hat >> > > _______________________________________________ > infinispan-dev mailing list > infinispan-dev(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/infinispan-dev _______________________________________________ infinispan-dev mailing list infinispan-dev(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/infinispan-dev

> On 22 Sep 2017, at 17:58, Sebastian Laskawiec <slaskawi(a)redhat.com&gt; wrote: > > > > On Fri, Sep 22, 2017 at 5:05 PM Sanne Grinovero <sanne(a)infinispan.org&gt; wrote: > On 22 September 2017 at 13:49, Sebastian Laskawiec <slaskawi(a)redhat.com&gt; wrote: > > It's very tricky... > > > > Memory is adjusted automatically to the container size [1] (of course you > > may override it by supplying Xmx or "-n" as parameters [2]). The safe limit > > is roughly Xmx=Xms=50% of container capacity (unless you do the off-heap, > > that you can squeeze Infinispan much, much more). > > > > Then there are Limits, Requests and QoS in Kubernetes [3][4]. We are in > > bustable memory category so if there is additional memory in the node, we'll > > get it. But if not, we won't go below 512 MB (and 500 mCPU). > > I hope that's a temporary choice of the work in process? > > Doesn't sound acceptable to address real world requirements.. > Infinispan expects users to estimate how much memory they will need - > which is hard enough - and then we should at least be able to start a > cluster to address the specified need. Being able to rely on 512MB > only per node would require lots of nodes even for small data sets, > leading to extreme resource waste as each node would consume some non > negligible portion of memory just to run the thing. > > hmmm yeah - its finished. > > I'm not exactly sure where the problem is. Is it 512 MB RAM/500 mCPUs? Or setting 50% of container memory? > > If the former and you set nothing, you will get the worse QoS and Kubernetes will shut your container in first order whenever it gets out of resources (I really recommend reading [4] and watching [3]). If the latter, yeah I guess we can tune it a little with off-heap but, as my the latest tests showed, if you enable RocksDB Cache Store, allocating even 50% is too much (the container got killed by OOM Killer). That's probably the reason why setting MaxRAM JVM parameters sets Xmx to 25% (!!!) of MaxRAM value. So even setting it to 50% means that we take the risk... > > So TBH, I see no silver bullet here and I'm open for suggestions. IMO if you're really know what you're doing, you should set Xmx yourself (this will turn off setting Xmx automatically by the bootstrap script) and possibly set limits (and adjust requests) in your Deployment Configuration (if you set both requests and limits you will have the best QoS). Try put it this way: I've just started an Infinispan ephermeral instance and trying to load some data and it's running out of memory. What knobs/settings does the template offer to make sure I have a big enough Infinispan instance(s) to handle my data?

(Don't reply with: make your data smaller) Cheers, > > > Thanks, > Sanne > > > > > Thanks, > > Sebastian > > > > [1] > > https://github.com/jboss-dockerfiles/infinispan/tree/master/server#adjust... > > [2] > > https://github.com/jboss-dockerfiles/infinispan/blob/master/server/docker... > > [3] https://www.youtube.com/watch?v=nWGkvrIPqJ4 > > [4] > > https://docs.openshift.com/enterprise/3.2/dev_guide/compute_resources.html > > > > On Fri, Sep 22, 2017 at 2:33 PM Galder Zamarreño <galder(a)redhat.com&gt; wrote: > >> > >> Hi Sebastian, > >> > >> How do you change memory settings for Infinispan started via service > >> catalog? > >> > >> The memory settings seem defined in [1], but this is not one of the > >> parameters supported. > >> > >> I guess we want this as parameter? > >> > >> Cheers, > >> > >> [1] > >> https://github.com/infinispan/infinispan-openshift-templates/blob/master/... > >> -- > >> Galder Zamarreño > >> Infinispan, Red Hat > >> > > > > _______________________________________________ > > infinispan-dev mailing list > > infinispan-dev(a)lists.jboss.org > > https://lists.jboss.org/mailman/listinfo/infinispan-dev > > _______________________________________________ > infinispan-dev mailing list > infinispan-dev(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/infinispan-dev > _______________________________________________ > infinispan-dev mailing list > infinispan-dev(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/infinispan-dev -- Galder Zamarreño Infinispan, Red Hat _______________________________________________ infinispan-dev mailing list infinispan-dev(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/infinispan-dev

On Fri, Sep 22, 2017 at 5:05 PM Sanne Grinovero <sanne(a)infinispan.org&gt; wrote: > > On 22 September 2017 at 13:49, Sebastian Laskawiec <slaskawi(a)redhat.com&gt; > wrote: > > It's very tricky... > > > > Memory is adjusted automatically to the container size [1] (of course > > you > > may override it by supplying Xmx or "-n" as parameters [2]). The safe > > limit > > is roughly Xmx=Xms=50% of container capacity (unless you do the > > off-heap, > > that you can squeeze Infinispan much, much more). > > > > Then there are Limits, Requests and QoS in Kubernetes [3][4]. We are in > > bustable memory category so if there is additional memory in the node, > > we'll > > get it. But if not, we won't go below 512 MB (and 500 mCPU). > > I hope that's a temporary choice of the work in process? > > Doesn't sound acceptable to address real world requirements.. > Infinispan expects users to estimate how much memory they will need - > which is hard enough - and then we should at least be able to start a > cluster to address the specified need. Being able to rely on 512MB > only per node would require lots of nodes even for small data sets, > leading to extreme resource waste as each node would consume some non > negligible portion of memory just to run the thing. hmmm yeah - its finished. I'm not exactly sure where the problem is. Is it 512 MB RAM/500 mCPUs? Or setting 50% of container memory?

If the former and you set nothing, you will get the worse QoS and Kubernetes will shut your container in first order whenever it gets out of resources (I really recommend reading [4] and watching [3]). If the latter, yeah I guess we can tune it a little with off-heap but, as my the latest tests showed, if you enable RocksDB Cache Store, allocating even 50% is too much (the container got killed by OOM Killer). That's probably the reason why setting MaxRAM JVM parameters sets Xmx to 25% (!!!) of MaxRAM value. So even setting it to 50% means that we take the risk... So TBH, I see no silver bullet here and I'm open for suggestions. IMO if you're really know what you're doing, you should set Xmx yourself (this will turn off setting Xmx automatically by the bootstrap script) and possibly set limits (and adjust requests) in your Deployment Configuration (if you set both requests and limits you will have the best QoS).

> > Thanks, > Sanne > > > > > Thanks, > > Sebastian > > > > [1] > > > > https://github.com/jboss-dockerfiles/infinispan/tree/master/server#adjust... > > [2] > > > > https://github.com/jboss-dockerfiles/infinispan/blob/master/server/docker... > > [3] https://www.youtube.com/watch?v=nWGkvrIPqJ4 > > [4] > > > > https://docs.openshift.com/enterprise/3.2/dev_guide/compute_resources.html > > > > On Fri, Sep 22, 2017 at 2:33 PM Galder Zamarreño <galder(a)redhat.com&gt; > > wrote: > >> > >> Hi Sebastian, > >> > >> How do you change memory settings for Infinispan started via service > >> catalog? > >> > >> The memory settings seem defined in [1], but this is not one of the > >> parameters supported. > >> > >> I guess we want this as parameter? > >> > >> Cheers, > >> > >> [1] > >> > >> https://github.com/infinispan/infinispan-openshift-templates/blob/master/... > >> -- > >> Galder Zamarreño > >> Infinispan, Red Hat > >> > > > > _______________________________________________ > > infinispan-dev mailing list > > infinispan-dev(a)lists.jboss.org > > https://lists.jboss.org/mailman/listinfo/infinispan-dev > > _______________________________________________ > infinispan-dev mailing list > infinispan-dev(a)lists.jboss.org > https://lists.jboss.org/mailman/listinfo/infinispan-dev _______________________________________________ infinispan-dev mailing list infinispan-dev(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/infinispan-dev

I don't understand your reply here... are you talking about Infinispan instances deployed on OpenShift Online? Or on premise?

I can understand having some limits for OpenShift Online, but these templates should also be applicable on premise, in which case I should be able to easily define how much memory I want for the data grid, and the rest of the parameters would be worked out by OpenShift/Kubernetes?

To demand on premise users to go and change their template just to adjust the memory settings seems to me goes against all the usability improvements we're trying to achieve.

Cheers, > On 22 Sep 2017, at 14:49, Sebastian Laskawiec <slaskawi(a)redhat.com&gt; wrote: > > It's very tricky... > > Memory is adjusted automatically to the container size [1] (of course you may override it by supplying Xmx or "-n" as parameters [2]). The safe limit is roughly Xmx=Xms=50% of container capacity (unless you do the off-heap, that you can squeeze Infinispan much, much more). > > Then there are Limits, Requests and QoS in Kubernetes [3][4]. We are in bustable memory category so if there is additional memory in the node, we'll get it. But if not, we won't go below 512 MB (and 500 mCPU). > > Thanks, > Sebastian > > [1] https://github.com/jboss-dockerfiles/infinispan/tree/master/server#adjust... > [2] https://github.com/jboss-dockerfiles/infinispan/blob/master/server/docker... > [3] https://www.youtube.com/watch?v=nWGkvrIPqJ4 > [4] https://docs.openshift.com/enterprise/3.2/dev_guide/compute_resources.html > > On Fri, Sep 22, 2017 at 2:33 PM Galder Zamarreño <galder(a)redhat.com&gt; wrote: > Hi Sebastian, > > How do you change memory settings for Infinispan started via service catalog? > > The memory settings seem defined in [1], but this is not one of the parameters supported. > > I guess we want this as parameter? > > Cheers, > > [1] https://github.com/infinispan/infinispan-openshift-templates/blob/master/... > -- > Galder Zamarreño > Infinispan, Red Hat > -- Galder Zamarreño Infinispan, Red Hat

Sebastian, What Galder, Sanne and others are saying is that in OpenShift on prem, there is no or at least a higher limit in the minimal container memory you can ask. And in these deployment, Infinispan should target the multi GB, not 512 MB. Of course, *if* you ask for a guaranteed 512MB, then it would be silly to try and consume more. On 25 Sep 2017, at 12:30, Sebastian Laskawiec <slaskawi(a)redhat.com&gt; wrote: On Mon, Sep 25, 2017 at 11:54 AM Galder Zamarreño <galder(a)redhat.com&gt; wrote: > I don't understand your reply here... are you talking about Infinispan > instances deployed on OpenShift Online? Or on premise? > TBH - I think there is no difference, so I'm thinking about both. > I can understand having some limits for OpenShift Online, but these > templates should also be applicable on premise, in which case I should be > able to easily define how much memory I want for the data grid, and the > rest of the parameters would be worked out by OpenShift/Kubernetes? > I have written a couple of emails about this on internal mailing list. Let me just point of some bits here: - We need to set either Xmx or MaxRAM to tell the JVM how much memory it can allocate. As you probably know JDK8 is not CGroups aware by default (there are some experimental options but they set MaxRAM parameter equal to CGroups limit; this translates to Xmx=MaxRAM(CGroups limit) / 4. I guess allocating Xmx=(CGroups limit)/4 is too high for us, so we need to set it explicitly. - in our Docker image we set Xmx = 50% of CGroups limit. This is better than settings above but there is some risk in certain scenarios. - As I mentioned in my previous email, in the templates we are setting Requests (not Limits!!!). So you will probably get more memory than specified in the template but it depends on the node you're running on. The key point is that you won't get less than those 512 MB. - You can always edit your DeploymentConfig (after creating your application from template) and adjust Limits (or even requests). - For simple scenarios and bigger containers (like 4 GB) we can go more than 50% (see internal mailing list for details). And as I said before - if you guys think we should do it differently, I'm open for suggestions. I think it's quite standard way of configuring this sort of stuff. > > To demand on premise users to go and change their template just to adjust > the memory settings seems to me goes against all the usability improvements > we're trying to achieve. > At some point you need to define how much memory you will need. Whether it's in the template, your DeploymentConfiguration (created from template using oc process), Quota - it doesn't matter. You must write it somewhere - don't you? With current approach, the best way to do it is in Deployment Configuration Requests. This sets CGroups limit, and based on that, Infinispan bootstrap scripts will calculate Xmx. > > Cheers, > > > On 22 Sep 2017, at 14:49, Sebastian Laskawiec <slaskawi(a)redhat.com&gt; > wrote: > > > > It's very tricky... > > > > Memory is adjusted automatically to the container size [1] (of course > you may override it by supplying Xmx or "-n" as parameters [2]). The safe > limit is roughly Xmx=Xms=50% of container capacity (unless you do the > off-heap, that you can squeeze Infinispan much, much more). > > > > Then there are Limits, Requests and QoS in Kubernetes [3][4]. We are in > bustable memory category so if there is additional memory in the node, > we'll get it. But if not, we won't go below 512 MB (and 500 mCPU). > > > > Thanks, > > Sebastian > > > > [1] > https://github.com/jboss-dockerfiles/infinispan/tree/master/server#adjust... > > [2] > https://github.com/jboss-dockerfiles/infinispan/blob/master/server/docker... > > [3] https://www.youtube.com/watch?v=nWGkvrIPqJ4 > > [4] > https://docs.openshift.com/enterprise/3.2/dev_guide/compute_resources.html > > > > On Fri, Sep 22, 2017 at 2:33 PM Galder Zamarreño <galder(a)redhat.com&gt; > wrote: > > Hi Sebastian, > > > > How do you change memory settings for Infinispan started via service > catalog? > > > > The memory settings seem defined in [1], but this is not one of the > parameters supported. > > > > I guess we want this as parameter? > > > > Cheers, > > > > [1] > https://github.com/infinispan/infinispan-openshift-templates/blob/master/... > > -- > > Galder Zamarreño > > Infinispan, Red Hat > > > > -- > Galder Zamarreño > Infinispan, Red Hat > > _______________________________________________ infinispan-dev mailing list infinispan-dev(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/infinispan-dev _______________________________________________ infinispan-dev mailing list infinispan-dev(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/infinispan-dev

On 28 Sep 2017, at 15:13, Emmanuel Bernard <emmanuel(a)hibernate.org&gt; wrote: I am personally content if you provide the total amount of memory for Pod and you as OSB designer decide of the -Xms/Xmx for the services. Unlike what Sanne said I think, Amazon and the like they don’t give you x GB of cache. They give you an instance of Redis or Memcached within a VM that has x amount of GB allocated. What you can stuck in is left as an exercise for the reader. Not ideal but I think they went for the practical in this case. For the pain JDG, then more options is fine. > On 28 Sep 2017, at 12:00, Sebastian Laskawiec <slaskawi(a)redhat.com <mailto:slaskawi@redhat.com>> wrote: > > So how about exposing two parameters - Xms/Xmx and Total amount of memory for Pod (Request = Limit in that case). Would it work for you? > > On Thu, Sep 28, 2017 at 8:38 AM Emmanuel Bernard <emmanuel(a)hibernate.org <mailto:emmanuel@hibernate.org>> wrote: > Sebastian, > > What Galder, Sanne and others are saying is that in OpenShift on prem, there is no or at least a higher limit in the minimal container memory you can ask. And in these deployment, Infinispan should target the multi GB, not 512 MB. > > Of course, *if* you ask for a guaranteed 512MB, then it would be silly to try and consume more. > > >> On 25 Sep 2017, at 12:30, Sebastian Laskawiec <slaskawi(a)redhat.com <mailto:slaskawi@redhat.com>> wrote: >> > >> >> >> On Mon, Sep 25, 2017 at 11:54 AM Galder Zamarreño <galder(a)redhat.com <mailto:galder@redhat.com>> wrote: >> I don't understand your reply here... are you talking about Infinispan instances deployed on OpenShift Online? Or on premise? >> >> TBH - I think there is no difference, so I'm thinking about both. >> >> I can understand having some limits for OpenShift Online, but these templates should also be applicable on premise, in which case I should be able to easily define how much memory I want for the data grid, and the rest of the parameters would be worked out by OpenShift/Kubernetes? >> >> I have written a couple of emails about this on internal mailing list. Let me just point of some bits here: >> We need to set either Xmx or MaxRAM to tell the JVM how much memory it can allocate. As you probably know JDK8 is not CGroups aware by default (there are some experimental options but they set MaxRAM parameter equal to CGroups limit; this translates to Xmx=MaxRAM(CGroups limit) / 4. I guess allocating Xmx=(CGroups limit)/4 is too high for us, so we need to set it explicitly. >> in our Docker image we set Xmx = 50% of CGroups limit. This is better than settings above but there is some risk in certain scenarios. >> As I mentioned in my previous email, in the templates we are setting Requests (not Limits!!!). So you will probably get more memory than specified in the template but it depends on the node you're running on. The key point is that you won't get less than those 512 MB. >> You can always edit your DeploymentConfig (after creating your application from template) and adjust Limits (or even requests). >> For simple scenarios and bigger containers (like 4 GB) we can go more than 50% (see internal mailing list for details). >> And as I said before - if you guys think we should do it differently, I'm open for suggestions. I think it's quite standard way of configuring this sort of stuff. >> >> To demand on premise users to go and change their template just to adjust the memory settings seems to me goes against all the usability improvements we're trying to achieve. >> >> At some point you need to define how much memory you will need. Whether it's in the template, your DeploymentConfiguration (created from template using oc process), Quota - it doesn't matter. You must write it somewhere - don't you? With current approach, the best way to do it is in Deployment Configuration Requests. This sets CGroups limit, and based on that, Infinispan bootstrap scripts will calculate Xmx. >> >> >> Cheers, >> >> > On 22 Sep 2017, at 14:49, Sebastian Laskawiec <slaskawi(a)redhat.com <mailto:slaskawi@redhat.com>> wrote: >> > >> > It's very tricky... >> > >> > Memory is adjusted automatically to the container size [1] (of course you may override it by supplying Xmx or "-n" as parameters [2]). The safe limit is roughly Xmx=Xms=50% of container capacity (unless you do the off-heap, that you can squeeze Infinispan much, much more). >> > >> > Then there are Limits, Requests and QoS in Kubernetes [3][4]. We are in bustable memory category so if there is additional memory in the node, we'll get it. But if not, we won't go below 512 MB (and 500 mCPU). >> > >> > Thanks, >> > Sebastian >> > >> > [1] https://github.com/jboss-dockerfiles/infinispan/tree/master/server#adjust... <https://github.com/jboss-dockerfiles/infinispan/tree/master/server#adjust... >> > [2] https://github.com/jboss-dockerfiles/infinispan/blob/master/server/docker... <https://github.com/jboss-dockerfiles/infinispan/blob/master/server/docker... >> > [3] https://www.youtube.com/watch?v=nWGkvrIPqJ4 <https://www.youtube.com/watch?v=nWGkvrIPqJ4> >> > [4] https://docs.openshift.com/enterprise/3.2/dev_guide/compute_resources.html <https://docs.openshift.com/enterprise/3.2/dev_guide/compute_resources.htm... >> > >> > On Fri, Sep 22, 2017 at 2:33 PM Galder Zamarreño <galder(a)redhat.com <mailto:galder@redhat.com>> wrote: >> > Hi Sebastian, >> > >> > How do you change memory settings for Infinispan started via service catalog? >> > >> > The memory settings seem defined in [1], but this is not one of the parameters supported. >> > >> > I guess we want this as parameter? >> > >> > Cheers, >> > >> > [1] https://github.com/infinispan/infinispan-openshift-templates/blob/master/... <https://github.com/infinispan/infinispan-openshift-templates/blob/master/... >> > -- >> > Galder Zamarreño >> > Infinispan, Red Hat >> > >> >> -- >> Galder Zamarreño >> Infinispan, Red Hat >> > >> _______________________________________________ >> infinispan-dev mailing list >> infinispan-dev(a)lists.jboss.org <mailto:infinispan-dev@lists.jboss.org> >> https://lists.jboss.org/mailman/listinfo/infinispan-dev <https://lists.jboss.org/mailman/listinfo/infinispan-dev>______________... > infinispan-dev mailing list > infinispan-dev(a)lists.jboss.org <mailto:infinispan-dev@lists.jboss.org> > https://lists.jboss.org/mailman/listinfo/infinispan-dev <https://lists.jboss.org/mailman/listinfo/infinispan-dev>______________... > infinispan-dev mailing list > infinispan-dev(a)lists.jboss.org <mailto:infinispan-dev@lists.jboss.org> > https://lists.jboss.org/mailman/listinfo/infinispan-dev _______________________________________________ infinispan-dev mailing list infinispan-dev(a)lists.jboss.org https://lists.jboss.org/mailman/listinfo/infinispan-dev