In Kubernetes, I have a statefulset with a number of replicas.
I've set the updateStrategy to RollingUpdate.
I've set podManagementPolicy to Parallel.
My statefulset instances do not have a persistent volume claim -- I use the statefulset as a way to allocate ordinals 0..(N-1) to pods in a deterministic manner.
The main reason for this, is to keep availability for new requests while rolling out software updates (freshly built containers) while still allowing each container, and other services in the cluster, to "know" its ordinal.
The behavior I want, when doing a rolling update, is for the previous statefulset pods to linger while there are still long-running requests processing on them, but I want new traffic to go to the new pods in the statefulset (mapped by the ordinal) without a temporary outage.
Unfortunately, I don't see a way of doing this -- what am I missing?
Because I don't use volume claims, you might think I could use deployments instead, but I really do need each of the pods to have a deterministic ordinal, that:
is unique at the point of dispatching new service requests (incoming HTTP requests, including public ingresses)
is discoverable by the pod itself
is persistent for the duration of the pod lifetime
is contiguous from 0 .. (N-1)
The second-best option I can think of is using something like zookeeper or etcd to separately manage this property, using some of the traditional long-poll or leader-election mechanisms, but given that kubernetes already knows (or can know) about all the necessary bits, AND kubernetes service mapping knows how to steer incoming requests from old instances to new instances, that seems more redundant and complicated than necessary, so I'd like to avoid that.
I assume that you need this for a stateful workload, a workload that e.g. requires writes. Otherwise you can use Deployments with multiple pods online for your shards. A key feature with StatefulSet is that they provide unique stable network identities for the instances.
The behavior I want, when doing a rolling update, is for the previous statefulset pods to linger while there are still long-running requests processing on them, but I want new traffic to go to the new pods in the statefulset.
This behavior is supported by Kubernetes pods. But you also need to implement support for it in your application.
New traffic will not be sent to your "old" pods.
A SIGTERM signal will be sent to the pod - your application may want to listen to this and do some action.
After a configurable "termination grace period", your pod will get killed.
See Kubernetes best practices: terminating with grace for more info about pod termination.
Be aware that you should connect to services instead of directly to pods for this to work. E.g. you need to create headless services for the replicas in a StatefulSet.
If your clients are connecting to a specific headless service, e.g. N, this means that it will not be available for some times during upgrades. You need to decide if your clients should retry their connections during this time period or if they should connect to another headless service if N is not available.
If you are in a case where you need:
stateful workload (e.g. support for write operations)
want high availability for your instances
then you need a form of distributed system that does some form of replication/synchronization, e.g. using raft or a product that implements this. Such system is easiest deployed as a StatefulSet.
You may be able to do this using Container Lifecycle Hooks, specifically the preStop hook.
We use this to drain connections from our Varnish service before it terminates.
However, you would need to implement (or find) a script to do the draining.
Related
I know why use StatefulSet for stateful applications. (e.g. DB or something)
In most cases, I can see like "You want to deploy stateful app to k8s? Use StatefulSet!"
However, I couldn't see like "You want to deploy stateless app to k8s? Then, DO NOT USE StatefulSet" ever.
Even nobody says "I don't recommend to use StatefulSet for stateless app", many stateless apps is deployed through Deployment, like it is the standard.
The StatefulSet has clear pros for stateful app, but I think Deployment doesn't for stateless app.
Is there any pros in Deployment for stateless apps? Or is there any clear cons in StatefulSet for stateless apps?
I supposed that StatefulSet cannot use LoadBalancer Service or StatefulSet has penalty to use HPA, but all these are wrong.
I'm really curious about this question.
P.S. Precondition is the stateless app also uses the PV, but not persists stateful data, for example logs.
I googled "When not to use StatefulSet", "when Deployment is better than StatefulSet", "Why Deployment is used for stateless apps", or something more questions.
I also see the k8s docs about StatefulSet either.
Different Priorities
What happens when a Node becomes unreachable in a cluster?
Deployment - Stateless apps
You want to maximize availability. As soon as Kubernetes detects that there are fewer than the desired number of replicas running in your cluster, the controllers spawn new replicas of it. Since these apps are stateless, it is very easy to do for the Kubernetes controllers.
StatefulSet - Stateful apps
You want to maximize availability - but not you must ensure data consistency (the state). To ensure data consistency, each replica has its own unique ID, and there are never multiple replicas of this ID, e.g. it is unique. This means that you cannot spawn up a new replica, unless that you are sure that the replica on the unreachable Node are terminated (e.g. stops using the Persistent Volume).
Conclusion
Both Deployment and StatefulSet try to maximize the availability - but StatefulSet cannot sacrifice data consistency (e.g. your state), so it cannot act as fast as Deployment (stateless) apps can.
These priorities does not only happens when a Node becomes unreachable, but at all times, e.g. also during upgrades and deployments.
In contrast to a Kubernetes Deployment, where pods are easily replaceable, each pod in a StatefulSet is given a name and treated individually. Pods with distinct identities are necessary for stateful applications.
This implies that if any pod perishes, it will be apparent right away. StatefulSets act as controllers but do not generate ReplicaSets; rather, they generate pods with distinctive names that follow a predefined pattern. The ordinal index appears in the DNS name of a pod. A distinct persistent volume claim (PVC) is created for each pod, and each replica in a StatefulSet has its own state.
For instance, a StatefulSet with four replicas generates four pods, each of which has its own volume, or four PVCs. StatefulSets require a headless service to return the IPs of the associated pods and enable direct interaction with them. The headless service has a service IP but no IP address and has to be created separately.The major components of a StatefulSet are the set itself, the persistent volume and the headless service.
That all being said, people deploy Stateful Applications with Deployments, usually they mount a RWX PV into the pods so all "frontends" share the same backend. Quite common in CNCF projects.
A stateful set manages each POD with a unique hostname based on an index number. So with an index, it would be easy to identify the individual PODs and also easy for the application to check which on rely or unique network identities. Also, you might have read stateful sets get deleted in a specified order to maintain consistency.
When you use stateful for the stateless application it will be like a burden to manage and add complexity to unique network identities and ordering guarantees.
For example, when you scale down to zero stateful sets it goes in the controlled way while with deployment or RS it won't be the same case. However, there is no guarantee when deleting the resource stateful set.
Also, Before a scaling operation is applied to a stateful set Pod, all of its predecessors must be Running and Ready. So if you are deploying the application, three Pods will be deployed suppose in order app-0, app-1, app-2. app-1 wont be deployed before app-0 is Running & Ready, and app-2 wont be deployed until app-1 is Ready.
While with deployment you can manage the % for and handle the RollingUpdate scenario but with a stateful set it will delete and recreate new POD one by one.
We have an AKS cluster and we want to achieve below two points in our architecture:
We have replicas of pods and we want to have only 1 request served by one pod. basically one pod - one request design.
When all pods are busy, then next coming request should not be queued at POD level, instead it should be queued at service level and once any of busy pod become idle or available then only queued request should be dispatched on idle pod.
How to achieve above things?
Generally, this could be achieved by creating a custom proxy that creates pods on demand, but in practice it will be very difficult and performance will be poor. This was very well explained by David Maze in his comment:
You need to write a custom proxy with access to the Kubernetes API that can create new pods on demand; this is not a standard Kubernetes setup. This is also an extremely heavy-weight setup (if it takes tens of seconds to pull and deploy a new pod you can hit HTTP request timeouts very easily) and every Web framework supports handling multiple requests per process.
I am having a stateful application - I am keeping data in user's sessions (basically data in HttpSession object) - but I do not have any requirement to write anything to persistent disk.
From what I have read so far - StatefulSet workloads are meant for stateful applications, but my understanding so far is that even though my application is a stateful application but Deployment workloads can also suffice my requirement because I do not want to write anything to persistent disks.
However, one point I am not sure about is that suppose I use Deployment workload and a lot of user data is present in my HttpSession object, now due to some reason Kubernetes restarts my Pod then of course all that user session data will be lost. So, my question are following:
Does StatefulSet handles this situation any better than Deployment workload?
So, only difference between Deployment workload and StatefulSet workload is about absence/presence of persistent disk or there is something to do with application session management as well in case of StatefulSet?
Does StatefulSet handles this situation any better than Deployment workload?
No. Neither Deployment nor StatefulSet will preserve memory contents. To preserve session information, you'll need to store it somewhere. One common approach is to use Redis.
So, only difference between Deployment workload and StatefulSet workload is about absence/presence of persistent disk or there is something to do with application session management as well in case of StatefulSet?
No, there are other differences:
StatefulSets create (and re-create) deterministic, consistent pod names (identifiers).
StatefulSets, are deployed, scaled, and updated one by one in a deterministic, consistent order. The next pod will be created only after the previous one reached the Running state.
Additionally, it's worth mentioning that persistent disks can be attached to pods that aren't part of a StatefulSet. It's just that it's convenient to have disks always be attached to a pod with a consistent id. For instance if you have pods running a replicated database, you can use StatefulSets to ensure that the master replica's disk is always attached to pod #1.
Edit:
Link to official documentation about StatefulSets
From the documentation:
Like a Deployment, a StatefulSet manages Pods that are based on an
identical container spec. Unlike a Deployment, a StatefulSet maintains
a sticky identity for each of their Pods. These pods are created from
the same spec, but are not interchangeable: each has a persistent
identifier that it maintains across any rescheduling.
...
StatefulSets are valuable for applications that require one or more of
the following.
Stable, unique network identifiers.
Stable, persistent storage.
Ordered, graceful deployment and scaling.
Ordered, automated rolling updates.
In the above, stable is synonymous with persistence across Pod
(re)scheduling. If an application doesn't require any stable
identifiers or ordered deployment, deletion, or scaling, you should
deploy your application using a workload object that provides a set of
stateless replicas. Deployment or ReplicaSet may be better suited to
your stateless needs.
Is there a way in Kubernetes to upgrade a given type of pod first when we have a deployment or stateful set with two or more replicas ( where one pod is master and others are not)?
My requirement to be specific is to ensure when calling upgrade on deployment/statefull set is to upgrade master as the last pod under a given number of replicas..
The only thing that's built into Kubernetes is the automatic sequential naming of StatefulSet pods.
If you have a StatefulSet, one of its pods is guaranteed to be named statefulsetname-0. That pod can declare itself the "master" for whatever purposes this is required. A pod can easily determine (by looking at its hostname(1)) whether it is the "master", and if it isn't, it can also easily determine what pod is. Updates happen by default in numerically reverse order, so statefulsetname-0 will be upgraded last, which matches your requirement.
StatefulSets have other properties, which you may or may not want. It's impossible for another pod to take over as the master if the first one fails; startup and shutdown happens in a fairly rigid order; if any part of your infrastructure is unstable then you may not be able to reliably scale the StatefulSet.
If you don't want a StatefulSet, you can implement your own leader election in a couple of ways (use a service like ZooKeeper or etcd that can help you maintain this state across the cluster; bring in a library for a leader-election algorithm like Raft). Kubernetes doesn't provide this on its own. The cluster will also be unaware of the "must upgrade the leader last" requirement, but if the current leader is terminated, another pod can take over the leadership.
The easiest way is probably having master in one deployment/statefulset, while followers in another deployment/statefulset. This approach ensure update is persist and can make use of update strategy in k8s.
The fact that k8s does not differentiate pod by containers nor any role specific to user application architecture ('master'); it is better to manage your own deployment when you have specific sequence that is outside of deployment/statefulset control. You can patch but change will not persist rollout restart.
I am trying to achieve the exactly-once delivery on Kafka using Spring-Kafka on Kubernetes.
As far as I understood, the transactional-ID must be set on the producer and it should be the same across restarts, as stated here https://stackoverflow.com/a/52304789/3467733.
The problem arises using this semantic on Kubernetes. How can you get a consistent ID?
To solve this problem I implementend a Spring boot application, let's call it "Replicas counter" that checks, through the Kubernetes API, how many pods there are with the same name as the caller, so I have a counter for every pod replica.
For example, suppose I want to deploy a Pod, let's call it APP-1.
This app does the following:
It perfoms a GET to the Replicas-Counter passing the pod-name as parameter.
The replicas-counter calls the Kubernetes API in order to check how many pods there are with that pod name. So it does a a +1 and returns it to the caller. I also need to count not-ready pods (think about a first deploy, they couldn't get the ID if I wasn't checking for not-ready pods).
The APP-1 gets the id and will use it as the transactional-id
But, as you can see a problem could arise when performing rolling updates, for example:
Suppose we have 3 pods:
At the beginning we have:
app-1: transactional-id-1
app-2: transactional-id-2
app-3: transactional-id-3
So, during a rolling update we would have:
old-app-1: transactional-id-1
old-app-2: transactional-id-2
old-app-3: transactional-id-3
new-app-3: transactional-id-4 (Not ready, waiting to be ready)
New-app-3 goes ready, so Kubernetes brings down the Old-app-3. So time to continue the rolling update.
old-app-1: transactional-id-1
old-app-2: transactional-id-2
new-app-3: transactional-id-4
new-app-2: transactional-id-4 (Not ready, waiting to be ready)
As you can see now I have 2 pods with the same transactional-id.
As far as I understood, these IDs have to be the same across restarts and unique.
How can I implement something that gives me consistent IDs? Is there someone that have dealt with this problem?
The problem with these IDs are only for the Kubernetes Deployments, not for the Stateful-Sets, as they have a stable identifier as name. I don't want to convert all deployment to stateful sets to solve this problem as I think it is not the correct way to handle this scenario.
The only way to guarantee the uniqueness of Pods is to use StatefulSet.
StatefulSets will allow you to keep the number of replicas alive but everytime pod dies it will be replaced with the same host and configuration. That will prevent data loss that is required.
Service in Statefulset must be headless because since each pod is going to be unique, so you are going to need certain traffic to reach certain pods.
Every pod require a PVC (in order to store data and recreate whenever pod is deleted from that data).
Here is a great article describing why StatefulSet should be used in similar case.