Courtesy : http://stevenpoitras.com/the-nutanix-bible/
The Nutanix Bible
Intro
Book of Nutanix
- Architecture
- How It Works
- Data Protection - updated!
- Data Locality
- Scalable Metadata
- Shadow Clones
- Elastic Dedupe Engine - updated!
- Networking and I/O
- CVM Autopathing
- Disk Balancing
- Software-Defined Controller Architecture
- Storage Tiering and Prioritization
- Storage Layers and Monitoring
- APIs & Interfaces
- Availability Domains - new!
- Data Protection - updated!
- Administration
- Troubleshooting
Book of vSphere
Book of Hyper-V
Revisions
Intro
Welcome to The Nutanix Bible! I work the with Nutanix platform on a daily basis – trying to find issues, push its limits as well as administer it for my production benchmarking lab. This page is being produced to serve as a living document outlining tips and tricks used every day by myself and a variety of engineers at Nutanix. This will also include summary items discussed as part of the Advanced Nutanix series. NOTE: This is not an official reference so tread at your own risk!Book of Nutanix
Architecture
Converged Platform
The Nutanix solution is a converged storage + compute solution which leverages local components and creates a distributed platform for virtualization aka virtual computing platform. The solution is a bundled hardware + software appliance which houses 2 (6000/7000 series) or 4 nodes (1000/2000/3000/3050 series) in a 2U footprint.Each node runs an industry standard hypervisor (ESXi, KVM, Hyper-V currently) and the Nutanix Controller VM (CVM). The Nutanix CVM is what runs the Nutanix software and serves all of the I/O operations for the hypervisor and all VMs running on that host. For the Nutanix units running VMware vSphere, the SCSI controller, which manages the SSD and HDD devices, is directly passed to the CVM leveraging VM-Direct Path (Intel VT-d). In the case of Hyper-V the storage devices are passed through to the CVM.
Below is an example of what a typical node logically looks like:
Together, a group of Nutanix Nodes forms a distributed platform called
the Nutanix Distributed Filesystem (NDFS). NDFS appears to the
hypervisor like any centralized storage array, however all of the I/Os
are handled locally to provide the highest performance. More detail on
how these nodes form a distributed system can be found below.Below is an example of how these Nutanix nodes form NDFS:
Cluster Components
The Nutanix platform is composed of the following high-level components:
Cassandra
- Key Role: Distributed metadata store
- Description: Medusa stores and manages all of the cluster metadata in a distributed ring like manner based upon a heavily modified Apache Cassandra. The Paxos algorithm is utilized to enforce strict consistency. This service runs on every node in the cluster. Cassandra is accessed via an interface called Medusa.
Zookeeper
- Key Role: Cluster configuration manager
- Description: Zeus stores all of the cluster configuration including hosts, IPs, state, etc. and is based upon Apache Zookeeper. This service runs on three nodes in the cluster, one of which is elected as a leader. The leader receives all requests and forwards them to the peers. If the leader fails to respond a new leader is automatically elected. Zookeeper is accessed via an interface called Zeus.
Stargate
- Key Role: Data I/O manager
- Description: Stargate is responsible for all data management and I/O operations and is the main interface from the hypervisor (via NFS, iSCSI or SMB). This service runs on every node in the cluster in order to serve localized I/O.
Curator
- Key Role: Map reduce cluster management and cleanup
- Description: Curator is responsible for managing and distributing tasks throughout the cluster including disk balancing, proactive scrubbing, and many more items. Curator runs on every node and is controlled by an elected Curator Master who is responsible for the task and job delegation.
Prism
- Key Role: UI and API
- Description: Prism is the management gateway for component and administrators to configure and monitor the Nutanix cluster. This includes Ncli, the HTML5 UI and REST API. Prism runs on every node in the cluster and uses an elected leader like all components in the cluster.
Genesis
- Key Role: Cluster component & service manager
- Description: Genesis is a process which runs on each node and is responsible for any services interactions (start/stop/etc.) as well as for the initial configuration. Genesis is a process which runs independently of the cluster and does not require the cluster to be configured/running. The only requirement for genesis to be running is that Zookeeper is up and running. The cluster_init and cluster_status pages are displayed by the genesis process.
Chronos
- Key Role: Job and Task scheduler
- Description: Chronos is responsible for taking the jobs and tasks resulting from a Curator scan and scheduling/throttling tasks among nodes. Chronos runs on every node and is controlled by an elected Chronos Master who is responsible for the task and job delegation and runs on the same node as the Curator Master.
Cerebro
- Key Role: Replication/DR manager
- Description: Cerebro is responsible for the replication and DR capabilities of NDFS. This includes the scheduling of snapshots, the replication to remote sites, and the site migration/failover. Cerebro runs on every node in the Nutanix cluster and all nodes participate in replication to remote clusters/sites.
Pithos
- Key Role: vDisk configuration manager
- Description: Pithos is responsible for vDisk (NDFS file) configuration data. Pithos runs on every node and is built on top of Cassandra.
Data Structure
The Nutanix Distributed Filesystem is composed of the following high-level structs:Storage Pool
- Key Role: Group of physical devices
- Description: A storage pool is a group of physical storage devices including PCIe SSD, SSD, and HDD devices for the cluster. The storage pool can span multiple Nutanix nodes and is expanded as the cluster scales. In most configurations only a single storage pool is leveraged.
Container
- Key Role: Group of VMs/files
- Description: A container is a logical segmentation of the Storage Pool and contains a group of VM or files (vDisks). Some configuration options (eg. RF) are configured at the container level, however are applied at the individual VM/file level. Containers typically have a 1 to 1 mapping with a datastore (in the case of NFS/SMB).
vDisk
- Key Role: vDisk
- Description: A vDisk is any file over 512KB on NDFS including .vmdks and VM hard disks. vDisks are composed of extents which are grouped and stored on disk as an extent group.
Extent
- Key Role: Logically contiguous data
- Description: A extent is a 1MB piece of logically contiguous data which consists of n number of contiguous blocks (varies depending on guest OS block size). Extents are written/read/modified on a sub-extent basis (aka slice) for granularity and efficiency. An extent’s slice may be trimmed when moving into the cache depending on the amount of data being read/cached.
Extent Group
- Key Role: Physically contiguous stored data
- Description: A extent group is a 1MB or 4MB piece of physically contiguous stored data. This data is stored as a file on the storage device owned by the CVM. Extents are dynamically distributed among extent groups to provide data striping across nodes/disks to improve performance. NOTE: as of 4.0 extent groups can now be either 1MB or 4MB depending on dedupe.
Here is another graphical representation of how these units are logically related:
I/O Path Overview
The Nutanix I/O path is composed of the following high-level components:
OpLog
- Key Role: Persistent write buffer
- Description: The Oplog is similar to a filesystem journal and is built to handle bursty writes, coalesce them and then sequentially drain the data to the extent store. Upon a write the OpLog is synchronously replicated to another n number of CVM’s OpLog before the write is acknowledged for data availability purposes. All CVM OpLogs partake in the replication and are dynamically chosen based upon load. The OpLog is stored on the SSD tier on the CVM to provide extremely fast write I/O performance, especially for random I/O workloads. For sequential workloads the OpLog is bypassed and the writes go directly to the extent store. If data is currently sitting in the OpLog and has not been drained, all read requests will be directly fulfilled from the OpLog until they have been drain where they would then be served by the extent store/content cache. For containers where fingerprinting (aka Dedupe) has been enabled, all write I/Os will be fingerprinted using a hashing scheme allowing them to be deduped based upon fingerprint in the content cache.
Extent Store
- Key Role: Persistent data storage
- Description: The Extent Store is the persistent bulk storage of NDFS and spans SSD and HDD and is extensible to facilitate additional devices/tiers. Data entering the extent store is either being A) drained from the OpLog or B) is sequential in nature and has bypassed the OpLog directly. Nutanix ILM will determine tier placement dynamically based upon I/O patterns and will move data between tiers.
Content Cache
- Key Role: Dynamic read cache
- Description: The Content Cache (aka “Elastic Dedupe Engine”) is a deduped read cache which spans both the CVM’s memory and SSD. Upon a read request of data not in the cache (or based upon a particular fingerprint) the data will be placed in to the single-touch pool of the content cache which completely sits in memory where it will use LRU until it is ejected from the cache. Any subsequent read request will “move” (no data is actually moved, just cache metadata) the data into the memory portion of the multi-touch pool which consists of both memory and SSD. From here there are two LRU cycles, one for the in-memory piece upon which eviction will move the data to the SSD section of the multi-touch pool where a new LRU counter is assigned. Any read request for data in the multi-touch pool will cause the data to go to the peak of the multi-touch pool where it will be given a new LRU counter. Fingerprinting is configured at the container level and can be configured via the UI. By default fingerprinting is disabled.
- Below we show a high-level overview of the Content Cache:
Extent Cache
- Key Role: In-memory read cache
- Description: The Extent Cache is an in-memory read cache that is completely in the CVM’s memory. This will store non-fingerprinted extents for containers where fingerprinting and dedupe is disabled. As of version 3.5 this is separate from the Content Cache, however these will be merging in a subsequent release.
How It Works
Data Protection
The Nutanix platform currently uses a resiliency factor aka replication factor (RF) and checksum to ensure data redundancy and availability in the case of a node or disk failure or corruption. As explained above the OpLog acts as a staging area to absorb incoming writes onto a low-latency SSD tier. Upon being written to the local OpLog the data is synchronously replicated to another one or two Nutanix CVM’s OpLog (dependent on RF) before being acknowledged (Ack) as a successful write to the host. This ensures that the data exists in at least two or three independent locations and is fault tolerant.NOTE: For RF3 a minimum of 5 nodes is required since metadata will be RF5. Data RF is configured via Prism and is done at the container level.
All nodes participate in OpLog replication to eliminate any “hot nodes” and ensuring linear performance at scale. While the data is being written a checksum is computed and stored as part of its metadata. Data is then asynchronously drained to the extent store where the RF is implicitly maintained. In the case of a node or disk failure the data is then re-replicated among all nodes in the cluster to maintain the RF. Any time the data is read the checksum is computed to ensure the data is valid. In the event where the checksum and data don’t match the replica of the data will be read and will replace the non-valid copy.
Below we show an example of what this logically looks like:
Data Locality
Being a converged (compute+storage) platform, I/O and data locality is key to cluster and VM performance with Nutanix. As explained above in the I/O path, all read/write IOs are served by the local Controller VM (CVM) which is on each hypervisor adjacent to normal VMs. A VM’s data is served locally from the CVM and sits on local disks under the CVM’s control. When a VM is moved from one hypervisor node to another (or during a HA event) the newly migrated VM’s data will be served by the now local CVM.When reading old data (stored on the now remote node/CVM) the I/O will be forwarded by the local CVM to the remote CVM. All write I/Os will occur locally right away. NDFS will detect the I/Os are occurring from a different node and will migrate the data locally in the background allowing for all read I/Os to now be served locally. The data will only be migrated on a read as to not flood the network.
Below we show an example of how data will “follow” the VM as it moves between hypervisor nodes:
Scalable Metadata
Metadata is at the core of any intelligent system and is even more critical for any filesystem or storage array. In terms of NDFS there are a few key structs that are critical for its success: it has to be right 100% of the time (aka. “strictly consistent”), it has to be scalable, and it has to perform, at massive scale. As mentioned in the architecture section above, NDFS utilizes a “ring like” structure as a key-value store which stores essential metadata as well as other platform data (eg. stats, etc.).In order to ensure metadata availability and redundancy a RF is utilized among an odd amount of nodes (eg. 3, 5, etc.). Upon a metadata write or update the row is written to a node in the ring and then replicated to n number of peers (where n is dependent on cluster size). A majority of nodes must agree before anything is committed which is enforced using the paxos algorigthm. This ensures strict consistency for all data and metadata stored as part of the platform.
Below we show an example of a metadata insert/update for a 4 node cluster:
Below we show an example of the metadata “ring” and how it scales:
Shadow Clones
The Nutanix Distributed Filesystem has a feature called ‘Shadow Clones’ which allows for distributed caching of particular vDisks or VM data which is in a ‘multi-reader’ scenario. A great example of this is during a VDI deployment many ‘linked clones’ will be forwarding read requests to a central master or ‘Base VM’. In the case of VMware View this is called the replica disk and is read by all linked clones and in XenDesktop this is called the MCS Master VM. This will also work in any scenario which may be a multi-reader scenario (eg. deployment servers, repositories, etc.).Data or I/O locality is critical for the highest possible VM performance and a key struct of NDFS. With Shadow Clones, NDFS will monitor vDisk access trends similar to what it does for data locality. However in the case there are requests occurring from more than two remote CVMs (as well as the local CVM), and all of the requests are read I/O, the vDisk will be marked as immutable. Once the disk has been marked as immutable the vDisk can then be cached locally by each CVM making read requests to it (aka Shadow Clones of the base vDisk). This will allow VMs on each node to read the Base VM’s vDisk locally.
In the case of VDI, this means the replica disk can be cached by each node and all read requests for the base will be served locally. NOTE: The data will only be migrated on a read as to not flood the network and allow for efficient cache utilization. In the case where the Base VM is modified the Shadow Clones will be dropped and the process will start over. Shadow clones are disabled by default (as of 3.5) and can be enabled/disabled using the following NCLI command: ncli cluster edit-params enable-shadow-clones=true.
Below we show an example of how Shadow Clones work and allow for distributed caching:
Elastic Dedupe Engine
The Elastic Dedupe Engine is a software based feature of NDFS which allows for data deduplication in the capacity (HDD) and performance (SSD/Memory) tiers. Sequential streams of data are fingerprinted during ingest using a SHA-1 hash at a 16K granularity. This fingerprint is only done on data ingest and is then stored persistently as part of the written block’s metadata. NOTE: Initially a 4K granularity was used for fingerprinting, however after testing 16K offered the best blend of dedupability with reduced metadata overhead. When deduped data is pulled into the cache this is done at 4K.Contrary to traditional approaches which utilize background scans, requiring the data to be re-read, Nutanix performs the fingerprint in-line on ingest. For duplicate data that can be deduplicated in the capacity tier the data does not need to be scanned or re-read, essentially duplicate copies can be removed.
Below we show an example of how the Elastic Dedupe Engine scales and handles local VM I/O requests:
Fingerprinting is done during data ingest of data with an I/O size of 64K or greater. Intel acceleration is leveraged for the SHA-1 computation which accounts for very minimal CPU overhead. In cases where fingerprinting is not done during ingest (eg. smaller I/O sizes), fingerprinting can be done as a background process. The Elastic Deduplication Engine spans both the capacity disk tier (HDD), but also the performance tier (SSD/Memory). As duplicate data is determined, based upon multiple copies of the same fingerprints, a background process will remove the duplicate data using the NDFS Map Reduce framework (curator).
For data that is being read, the data will be pulled into the NDFS Content Cache which is a multi-tier/pool cache. Any subsequent requests for data having the same fingerprint will be pulled directly from the cache. To learn more about the Content Cache and pool structure, please refer to the ‘Content Cache’ sub-section in the I/O path overview, or click HERE.
Below we show an example of how the Elastic Dedupe Engine interacts with the NDFS I/O path:
Networking and I/O
The Nutanix platform does not leverage any backplane for inter-node communication and only relies on a standard 10GbE network. All storage I/O for VMs running on a Nutanix node is handled by the hypervisor on a dedicated private network. The I/O request will be handled by the hypervisor which will then forward the request to the private IP on the local CVM. The CVM will then perform the remote replication with other Nutanix nodes using its external IP over the public 10GbE network.For all read requests these will be served completely locally in most cases and never touch the 10GbE network. This means that the only traffic touching the public 10GbE network will be NDFS remote replication traffic and VM network I/O. There will however be cases where the CVM will forward requests to other CVMs in the cluster in the case of a CVM being down or data being remote. Also, cluster wide tasks such as disk balancing will temporarily generate I/O on the 10GbE network.
Below we show an example of how the VM’s I/O path interacts with the private and public 10GbE network:
CVM Autopathing
Reliability and resiliency is a key, if not the most important, piece to NDFS. Being a distributed system NDFS is built to handle component, service and CVM failures. In this section I’ll cover how CVM “failures” are handled (I’ll cover how we handle component failures in future update). A CVM “failure” could include a user powering down the CVM, a CVM rolling upgrade, or any event which might bring down the CVM.NDFS has a feature called autopathing where when a local CVM becomes unavailable the I/Os are then transparently handled by other CVMs in the cluster. The hypervisor and CVM communicate using a private 192.168.5.0 network on a dedicated vSwitch (more on this above). This means that for all storage I/Os these are happening to the internal IP addresses on the CVM (192.168.5.2). The external IP address of the CVM is used for remote replication and for CVM communication.
Below we show an example of what this looks like:
In the event of a local CVM failure the local 192.168.5.2 addresses
previously hosted by the local CVM is unavailable. NDFS will
automatically detect this outage and will redirect these I/Os to another
CVM in the cluster over 10GbE. The re-routing is done transparently to
the hypervisor and VMs running on the host. This means that even if a
CVM is powered down the VMs will still continue to be able to perform
I/Os to NDFS. NDFS is also self-healing meaning it will detect the CVM
has been powered off and will automatically reboot or power-on the local
CVM. Once the local CVM is back up and available, traffic will then
seamlessly be transferred back and served by the local CVM.Below we show a graphical representation of how this looks for a failed CVM:
Disk Balancing
NDFS is designed to be a very dynamic platform which can react to various workloads as well as allow heterogeneous node types: compute heavy (3050, etc.) and storage heavy (60X0, etc.) to be mixed in a single cluster. Ensuring uniform distribution of data is an important item when mixing nodes with larger storage capacities.NDFS has a native feature called disk balancing which is used to ensure uniform distribution of data throughout the cluster. Disk balancing works on a node’s utilization of its local storage capacity and is integrated with NDFS ILM. Its goal is to keep utilization uniform among nodes once the utilization has breached a certain threshold.
Below we show an example of a mixed cluster (3050 + 6050) in a “unbalanced” state:
Disk balancing leverages the NDFS Curator framework and is run as a
scheduled process as well as when a threshold has been breached (eg.
local node capacity utilization > n %). In the case where the data
is not balanced Curator will determine which data needs to be moved and
will distribute the tasks to nodes in the cluster. In the case where the
node types are homogeneous (eg. 3050) utilization should be fairly
uniform.However, if there are certain VMs running on a node which are writing much more data than others there can become a skew in the per node capacity utilization. In this case disk balancing would run and move the coldest data on that node to other nodes in the cluster. In the case where the node types are heterogeneous (eg. 3050 + 6020/50/70), or where a node may be used in a “storage only” mode (not running any VMs), there will likely be a requirement to move data.
Below we show an example the mixed cluster after disk balancing has been run in a “balanced” state:
In some scenarios customers might run some nodes in a “storage only”
state where only the CVM will run on the node whose primary purpose is
bulk storage capacity. In this case the full nodes memory can be added
to the CVM to provide a much larger read cache.Below we show an example of how a storage only node would look in a mixed cluster with disk balancing moving data to it from the active VM nodes:
Software-Defined Controller Architecture
As mentioned above (likely numerous times), the Nutanix platform is a software based solution which ships as a bundled software + hardware appliance. The controller VM is where the vast majority of the Nutanix software and logic sits and was designed from the beginning to be an extensible and pluggable architecture.A key benefit to being software defined and not relying upon any hardware offloads or constructs is around extensibility. Like with any product life cycle there will always be advancements and new features which are introduced. By not relying on any custom ASIC/FPGA or hardware capabilities, Nutanix can develop and deploy these new features through a simple software update. This means that the deployment of a new feature (say deduplication) can be deployed by upgrading the current version of the Nutanix software. This also allows newer generation features to be deployed on legacy hardware models.
For example, say you’re running a workload running an older version of Nutanix software on a prior generation hardware platform (eg. 2400). The running software version doesn’t provide deduplication capabilities which your workload could benefit greatly from. To get these features you perform a rolling upgrade of the Nutanix software version while the workload is running, and whala you now have deduplication. It’s really that easy.
Similar to features, the ability to create new “adapters” or interfaces into NDFS is another key capability. When the product first shipped it solely supported iSCSI for I/O from the hypervisor, this has now grown to include NFS and SMB. In the future there is the ability to create new adapters for various workloads and hypervisors (HDFS, etc.). And again, all deployed via a software update.
This is contrary to mostly all legacy infrastructures as a hardware upgrade or software purchase was normally required to get the “latest and greatest” features. With Nutanix it’s different, since all features are deployed in software they can run on any hardware platform, any hypervisor and be deployed through simple software upgrades.
Below we show a logical representation of what this software-defined controller framework looks like:
Storage Tiering and Prioritization
The Disk Balancing section above talked about how storage capacity was pooled among all nodes in a Nutanix cluster and that ILM would be used to keep hot data local. A similar concept applies to disk tiering in which the cluster’s SSD and HDD tiers are cluster wide and NDFS ILM is responsible for triggering data movement events.A local node’s SSD tier is always the highest priority tier for all I/O generated by VMs running on that node, however all of the cluster’s SSD resources are made available to all nodes within the cluster. The SSD tier will always offer the highest performance and is a very important thing to manage for hybrid arrays.
The tier prioritization can be classified at a high-level by the following:
Specific types of resources (eg. SSD, HDD, etc.) are pooled together
and form a cluster wide storage tier. This means that any node within
the cluster can leverage the full tier capacity, regardless if it is
local or not.Below we show a high level example of how this pooled tiering looks:
A common question is what happens when a local node’s SSD becomes full?
As mentioned in the Disk Balancing section a key concept is trying to
keep uniform utilization of devices within disk tiers. In the case
where a local node’s SSD utilization is high, disk balancing will kick
in to move the coldest data on the local SSDs to the other SSDs
throughout the cluster. This will free up space on the local SSD to
allow the local node to write to SSD locally instead of going over the
network. A key point to mention is that all CVMs and SSDs are used for
this remote I/O to eliminate any potential bottlenecks and remediate
some of the hit by performing I/O over the network. 
The other case is when the overall tier utilization breaches a specific
threshold [curator_tier_usage_ilm_threshold_percent (Default=75)] where
NDFS ILM will kick in and as part of a Curator job will down-migrate
data from the SSD tier to the HDD tier. This will bring utilization
within the threshold mentioned above or free up space by the following
amount [curator_tier_free_up_percent_by_ilm (Default=15)], whichever is
greater.The data for down-migration is chosen using last access time. In the case where the SSD tier utilization is 95%, 20% of the data in the SSD tier will be moved to the HDD tier (95% –> 75%). However, if the utilization was 80% only 15% of the data would be moved to the HDD tier using the minimum tier free up amount.
NDFS ILM will constantly monitor the I/O patterns and (down/up)-migrate
data as necessary as well as bring the hottest data local regardless of
tier.Storage Layers and Monitoring
The Nutanix platform monitors storage at multiple layers throughout the stack ranging from the VM/Guest OS all the way down to the physical disk devices. Knowing the various tiers and how these relate is important whenever monitoring the solution and allows you to get full visibility of how the ops relate.Below we show the various layers of where operations are monitored and the relative granularity which are explained below:
Virtual Machine Layer
- Key Role: Metrics reported by the Guest OS
- Description: Virtual Machine or Guest OS level metrics are pulled directly from the hypervisor and represent the performance the Guest OS is seeing and is indicative of the I/O performance the application is seeing.
- When to use: When troubleshooting or looking for OS or application level detail
Hypervisor Layer
- Key Role: Metrics reported by the Hypervisor(s)
- Description: Hypervisor level metrics are pulled directly from the hypervisor and represent the most accurate metrics the hypervisor(s) are seeing. This data can be viewed for one of more hypervisor node(s) or the aggregate cluster. This layer will provide the most accurate data in terms of what performance the platform is seeing and should be leveraged in most cases. In certain scenarios the hypervisor may combine or split operations coming from VMs which can show the difference in metrics reported by the VM and hypervisor. These numbers will also include cache hits served by the Nutanix CVMs.
- When to use: Most common cases as this will provide the most detailed and valuable metrics
Controller Layer
- Key Role: Metrics reported by the Nutanix Controller(s)
- Description: Controller level metrics are pulled directly from the Nutanix Controller VMs (eg. Stargate 2009 page) and represent what the Nutanix front-end is seeing from NFS/SMB/iSCSI or any back-end operations (eg. ILM, disk balancing, etc.). This data can be viewed for one of more Controller VM(s) or the aggregate cluster. The metrics seen by the Controller Layer should match those seen by the hypervisor layer, however will include any backend operations (eg. ILM, disk balancing). These numbers will also include cache hits served by memory.
- When to use: Similar to the hypervisor layer, can be used to show how much backend operation is taking place
Disk Layer
- Key Role: Metrics reported by the Disk Device(s)
- Description: Disk level metrics are pulled directly from the physical disk devices (via the CVM) and represent what the back-end is seeing. This includes data hitting the OpLog or Extent Store where an I/O is performed on the disk. This data can be viewed for one of more disk(s), the disk(s) for a particular node or the aggregate disks in the cluster. In common cases it is expected that the disk ops should match the number of incoming writes as well as reads not served from the memory portion of the cache. Any reads being served by the memory portion of the cache will not be counted here as the op is not hitting the disk device.
- When to use: When looking to see how many ops are served from cache or hitting the disks
APIs and Interfaces
Core to any dynamic or “Software Defined” environment, Nutanix provides a vast array of interfaces allowing for simple programmability and interfacing. Here are the main interfaces:- REST API
- NCLI
- Scripting interfaces – more coming here soon
Below we show a small snippet of the Nutanix REST API explorer which allows developers to see the API and format:
Operations can be expanded to display details and examples of the REST call: 
Availability Domains
Availability Domains aka node/block/rack awareness is a key struct for distributed systems to abide by for determining component and data placement. NDFS is currently node and block aware, however this will increase to rack aware as cluster sizes grow. Nutanix refers to a “block” as the chassis which contains either one, two or four server “nodes”.NOTE: at minimum of 3 blocks must be utilized for block awareness to be activated, otherwise node awareness will be defaulted to. It is recommended to utilized uniformly populated blocks to ensure block awareness is enabled. Common scenarios and the awareness level utilized can be found at the bottom of this section. The 3 block requirement is due to ensure quorum.
For example a 3450 would be a block which holds 4 nodes. The reason for distributing roles or data across blocks to ensure if a block fails or needs maintenance the system can continue to run without interruption. NOTE: Within a block the redundant PSU and fans are the only shared components
Awareness can be broken into a few key focus areas:
- Data (The VM data)
- Metadata (Cassandra)
- Configuration Data (Zookeeper)
With NDFS data replicas will be written to other blocks in the cluster to ensure that in the case of a block failure or planned downtime, the data remains available. This is true for both RF2 and RF3 scenarios as well as in the case of a block failure.
An easy comparison would be “node awareness” where a replica would need to be replicated to another node which will provide protection in the case of a node failure. Block awareness further enhances this by providing data availability assurances in the case of block outages.
Below we show how the replica placement would work in a 3 block deployment:
In the case of a block failure, block awareness will be maintained and the re-replicated blocks will be replicated to other blocks within the cluster:
Metadata
As mentioned in the Scalable Metadata section above, Nutanix leverages a heavily modified Cassandra platform to store metadata and other essential information. Cassandra leverages a ring-like structure and replicates to n number of peers within the ring to ensure data consistency and availability.
Below we show an example of the Cassandra ring for a 12 node cluster:
Cassandra peer replication iterates through nodes in a clockwise manner throughout the ring. With block awareness the peers are distributed among the blocks to ensure no two peers are on the same block.
Below we show an example node layout translating the ring above into the block based layout:
With this block aware nature, in the event of a block failure there will still be at least two copies of the data (with Metadata RF3 – In larger clusters RF5 can be leveraged).
Below we show an example of all of the nodes replication topology to form the ring (yes – its a little busy):
Configuration Data
Nutanix leverages Zookeeper to store essential configuration data for the cluster. This role is also distributed in a block aware manner to ensure availability in the case of a block failure.
Below we show an example layout showing 3 Zookeeper nodes distributed in a block aware manner:
In the event of a block outage, meaning on of the Zookeeper nodes will be gone, the Zookeeper role would be transferred to another node in the cluster as shown below:
Below we breakdown some common scenarios and what level of awareness will be utilized:
- < 3 blocks –> NODE awareness
- 3+ blocks uniformly populated –> BLOCK + NODE awareness
- 3+ blocks not uniformly populated
- If SSD tier variance between blocks is > max variance –> NODE awareness
- Example: 2 x 3450 + 1 x 3150
- If SSD tier variance between blocks is < max variance –> BLOCK + NODE awareness
- Example: 2 x 3450 + 1 x 3350
- NOTE: max tier variance is calculated as: 100 / (RF+1)
- Eg. 33% for RF2 or 25% for RF3
- If SSD tier variance between blocks is > max variance –> NODE awareness
Administration
Important Pages
These are advanced Nutanix pages besides the standard user interface that allow you to monitor detailed stats and metrics. The URLs are formatted in the following way: http://<Nutanix CVM IP/DNS>:<Port/path (mentioned below)> Example: http://MyCVM-A:2009 NOTE: if you’re on a different subnet IPtables will need to be disabled on the CVM to access the pages.# 2009 Page
- This is a Stargate page used to monitor the back end storage system and should only be used by advanced users. I’ll have a post that explains the 2009 pages and things to look for.
# 2009/latency Page
- This is a Stargate page used to monitor the back end latency
# 2009/h/traces Page
- This is the Stargate page used to monitor activity traces for operations
# 2010 Page
- This is the Curator page which is used for monitoring curator runs
# 2010/master/control Page
- This is the Curator control page which is used to manually start Curator jobs
# 2011 Page
- This is the Chronos page which monitors jobs and tasks scheduled by curator
# 2020 Page
- This is the Cerebro page which monitors the protection domains, replication status and DR
# 7777 Page
- This is the Aegis Portal page which can be used to get good logs and statistics, useful commands, and modify Gflags
Cluster Commands
# Check cluster status
# Check local CVM service status
# Nutanix cluster upgrade
# Restart cluster service from CLI
# Start cluster service from CLI
# Restart local service from CLI
# Start local service from CLI
# Cluster add node from cmdline
# Find number of vDisks
# Find cluster id
# Disable IPtables
# Check for Shadow Clones
# Reset Latency Page Stats
# Find Number of vDisks
# Start Curator scan from CLI
# Compact ring
# Find NOS version
# Find CVM version
# Manually fingerprint vDisk(s)
# Echo Factory_Config.json for all cluster nodes
# Upgrade a single Nutanix node’s NOS version
# Install Nutanix Cluster Check (NCC)
# Run Nutanix Cluster Check (NCC)
NCLI
NOTE: All of these actions can be performed via the HTML5 GUI. I just use these commands as part of my bash scripting to automate tasks.# Add subnet to NFS whitelist
# Display Nutanix Version
# Display hidden NCLI options
# List Storage Pools
# List containers
# Create container
# List VMs
# List public keys
# Add public key
# Remove public key
# Create protection domain
# Create remote site
# Create protection domain for all VMs in container
# Create protection domain with specified VMs
# Create snapshot of protection domain
# Create snapshot and replication schedule to remote site
# List replication status
# Migrate protection domain to remote site
# Activate protection domain
# Enable NDFS Shadow Clones
# Enable Dedup for vDisk
Metrics & Thresholds
The below will cover specific metrics and thresholds on the Nutanix back end. More updates to these coming shortly!2009 Stargate – Overview
| Metric | Explanation | Threshold/Target |
| Start time | The start time of the Stargate service | |
| Build version | The build version currently running | |
| Build last commit date | The last commit date of the build | |
| Stargate handle | The Stargate handle | |
| iSCSI handle | The iSCSI handle | |
| SVM id | The SVM id of Stargate | |
| Incarnation id | ||
| Highest allocated opid | ||
| Highest contiguous completed opid | ||
| Extent cache hits | The % of read requests served directly from the in-memory extent cache | |
| Extent cache usage | The MB size of the extent cache | |
| Content cache hits | The % of read requests served directly from the content cache | |
| Content cache flash pagein pct | ||
| Content cache memory usage | The MB size of the in-memory content cache | |
| Content cache flash usage | The MB size of the SSD content cache | |
| QoS Queue (size/admitted) | The admission control queue size and number of admitted ops | |
| Oplog QoS queue (size/admitted) | The oplog queue size and number of admitted ops | |
| NFS Flush Queue (size/admitted) | ||
| NFS cache usage | ||
2009 Stargate – Cluster State
| Metric | Explanation | Threshold/Target |
| SVM Id | The Id of the Controller | |
| IP:port | The IP:port of the Stargate handle | |
| Incarnation | ||
| SSD-PCIe | The SSD-PCIe devices and size/utilization | |
| SSD-SATA | The SSD-SATA devices and size/utilization | |
| DAS-SATA | The HDD-SATA devices and size/utilization | |
| Container Id | The Id of the container | |
| Container Name | The Name of the container | |
| Max capacity (GB) – Storage pool | The Max capacity of the storage pool | |
| Max capacity (GB) – Container | The Max capacity of the container (will normally match the storage pool size) | |
| Reservation (GB) – Total across vdisks | The reservation in GB across vdisks | |
| Reservation (GB) – Admin provisioned | ||
| Container usage (GB) – Total | The total usage in GB per container | |
| Container usage (GB) – Reserved | The reservation used in GB per container | |
| Container usage (GB) – Garbage | ||
| Unreserved available (GB) – Container | The available capacity in GB per container | |
| Unreserved available (GB) – Storage pool | The available capacity in GB for the storage pool | |
2009 Stargate – NFS Slave
| Metric | Explanation | Threshold/Target |
| Vdisk Name | The name of the Vdisk on NDFS | |
| Unstable data – KB | ||
| Unstable data – Ops/s | ||
| Unstable data – KB/s | ||
| Outstanding Ops – Read | The number of outstanding read ops for the Vdisk | |
| Outstanding Ops – Write | The number of outstanding write ops for the Vdisk | |
| Ops/s – Read | The number of current read operations per second for the Vdisk | |
| Ops/s – Write | The number of current write operations per second for the Vdisk | |
| Ops/s – Error | The number of current error (failed) operations per second for the Vdisk | |
| KB/s – Read | The read throughput in KB/s for the Vdisk | |
| KB/s – Write | The write throughput in KB/s for the Vdisk | |
| Avg latency (usec) – Read | The average read op latecy in micro seconds for the Vdisk | |
| Avg latency (usec) – Write | The average write op latecy in micro seconds for the Vdisk | |
| Avg op size | The average op size in bytes for the Vdisk | |
| Avg outstanding | The average outstanding ops for the Vdisk | |
| % busy | The % busy of the Vdisk | |
| Container Name | The name of the container | |
| Outstanding Ops – Read | The number of outstanding read ops for the container | |
| Outstanding Ops – Write | The number of outstanding write ops for the container | |
| Outstanding Ops – NS lookup | The number of oustanding NFS lookup ops for the container | |
| Outstanding Ops – NS update | The number of outstanding NFS update ops for the container | |
| Ops/s – Read | The number of current read operations per second for the container | |
| Ops/s – Write | The number of current write operations per second for the container | |
| Ops/s – NS lookup | The number of current NFS lookup ops for the container | |
| Ops/s – NS update | The number of current NFS update ops for the container | |
| Ops/s – Error | The number of current error (failed) operations per second for the container | |
| KB/s – Read | The read throughput in KB/s for the container | |
| KB/s – Write | The write throughput in KB/s for the container | |
| Avg latency (usec) – Read | The average read op latecy in micro seconds for the container | |
| Avg latency (usec) – Write | The average write op latecy in micro seconds for the container | |
| Avg latency (usec) – NS lookup | The average NFS lookup latency in micro seconds for the container | |
| Avg latency (usec) – NS update | The average NFS lookup update in micro seconds for the container | |
| Avg op size | The average op size in bytes for the container | |
| Avg outstanding | The average outstanding ops for the container | |
| % busy | The % busy of the container | |
2009 Stargate – Hosted VDisks
| Metric | Explanation | Threshold/Target |
| Vdisk Id | The Id of the Vdisk on NDFS | |
| Vdisk Name | The name of the Vdisk on NDFS | |
| Usage (GB) | The usage in GB per Vdisk | |
| Dedup (GB) | ||
| Oplog – KB | The size of the Oplog for the Vdisk | |
| Oplog – Fragments | The number of fragments of the Oplog for the Vdisk | |
| Oplog – Ops/s | The number of current opeations per second for the Vdisk | |
| Oplog – KB/s | The throughput in KB/s for the Vdisk | |
| Outstanding Ops – Read | The number of outstanding read ops for the Vdisk | |
| Outstanding Ops – Write | The number of outstanding write ops for the Vdisk | |
| Outstanding Ops – Estore | The number of outstanding ops to the extent store for the Vdisk | |
| Ops/s – Read | The number of current read operations per second for the Vdisk | |
| Ops/s – Write | The number of current write operations per second for the Vdisk | |
| Ops/s – Error | The number of current error (failed) operations per second for the Vdisk | |
| Ops/s – Random | ||
| KB/s – Read | The read throughput in KB/s for the Vdisk | |
| KB/s – Write | The write throughput in KB/s for the Vdisk | |
| Avg latency (usec) | The average op latency in micro seconds for the Vdisk | |
| Avg op size | The average op size in bytes for the Vdisk | |
| Avg qlen | The average queue length for the Vdisk | |
| % busy | ||
2009 Stargate – Extent Store
| Metric | Explanation | Threshold/Target |
| Disk Id | The disk id of the physical device | |
| Mount point | The mount point of the physical device | |
| Outstanding Ops – QoS Queue | The number of (primary/secondary) ops for the device | |
| Outstanding Ops – Read | The number of outstanding read ops for the device | |
| Outstanding Ops – Write | The number of outstanding write ops for the device | |
| Outstanding Ops – Replicate | ||
| Outstanding Ops – Read Replica | ||
| Ops/s – Read | The number of current read operations per second for the device | |
| Ops/s – Write | The number of current write operations per second for the device | |
| Ops/s – Error | The number of current error (failed) operations per second for the device | |
| Ops/s – Random | ||
| KB/s – Read | The read throughput in KB/s for the device | |
| KB/s – Write | The write throughput in KB/s for the device | |
| Avg latency (usec) | The average op latency in micro seconds for the device | |
| Avg op size | The average op size in bytes for the device | |
| Avg qlen | The average queue length for the device | |
| Avg qdelay | The average queue delay for the device | |
| % busy | ||
| Size (GB) | ||
| Total usage (GB) | The total usage in GB for the device | |
| Unshared usage (GB) | ||
| Dedup usage (GB) | ||
| Garbage (GB) | ||
| Egroups | The number of extent groups for the device | |
| Corrupt Egroups | The number of corrupt (bad) extent groups for the device | |
Gflags
Coming soon
Troubleshooting
# Find cluster error logs
# Find cluster fatal logs
Book of vSphere
Architecture
To be inputHow It Works
Array Offloads – VAAI
The Nutanix platform supports the VMware APIs for Arry Integration (VAAI) which allows the hypervisor to offload certain tasks to the array. This is much more efficient as the hypervisor doesn’t need to be the “man in the middle”. Nutanix currently supports the VAAI primitives for NAS including the ‘full file clone’, ‘fast file clone’ and ‘reserve space’ primitives. Here’s a good article explaining the various primitives: LINK. For both the full and fast file clones a NDFS “fast clone” is done meaning a writable snapshot (using re-direct on write) for each clone is created. Each of these clones has its own block map meaning that chain depth isn’t anything to worry about. The following will determine whether or not VAAI will be used for specific scenarios:- Clone VM with Snapshot –> VAAI will NOT be used
- Clone VM without Snapshot which is Powered Off –> VAAI WILL be used
- Clone VM to a different Datastore/Container –> VAAI will NOT be used
- Clone VM which is Powered On –> VAAI will NOT be used
- View Full Clone (Template with Snapshot) –> VAAI will NOT be used
- View Full Clone (Template w/o Snapshot) –> VAAI WILL be used
- View Linked Clone (VCAI) –> VAAI WILL be used
Administration
To be inputImportant Pages
To be inputCommand Reference
# ESXi cluster upgrade
# Restart ESXi host services
# Display ESXi host nics in ‘Up’ state
# Display ESXi host 10GbE nics and status
# Display ESXi host active adapters
# Display ESXi host routing tables
# Check if VAAI is enabled on datastore
# Set VIB acceptance level to community supported
# Install VIB
# Check ESXi ramdisk space
# Clear pynfs logs
Metrics & Thresholds
To be inputTroubleshooting
To be inputBook of Hyper-V
Architecture
To be inputHow It Works
Array Offloads – ODX
The Nutanix platform supports the Microsoft Offloaded Data Transfers (ODX) which allows the hypervisor to offload certain tasks to the array. This is much more efficient as the hypervisor doesn’t need to be the “man in the middle”. Nutanix currently supports the ODX primitives for SMB which include full copy and zeroing operations. However contrary to VAAI which has a “fast file” clone operation (using writable snapshots) the ODX primitives do not have an equivalent and perform a full copy. Given this, it is more efficient to rely on the native NDFS clones which can currently be invoked via nCLI, REST, or Powershell CMDletsCurrently ODX IS invoked for the following operations:
- In VM or VM to VM file copy on NDFS SMB share
- SMB share file copy
- Deploy template from SCVMM Library (NDFS SMB share) - NOTE: Shares must be added to the SCVMM cluster using short names (eg. not FQDN). An easy way to force this is to add an entry into the hosts file for the cluster (eg. 10.10.10.10 nutanix-130).
- Clone VM through SCVMM
- Deploy template from SCVMM Library (non-NDFS SMB Share)
- XenDesktop Clone Deployment
Administration
To be inputImportant Pages
To be inputCommand Reference
# Execute command on multiple remote hosts
# Check available VMQ Offloads
# Disable VMQ for VMs matching a specific prefix
# Enable VMQ for VMs matching a certain prefix
# Power-On VMs matching a certain prefix
# Shutdown VMs matching a certain prefix
# Stop VMs matching a certain prefix
# Get Hyper-V host RSS settings
Metrics & Thresholds
To be inputTroubleshooting
To be inputRevisions
- 09-04-2013 | Initial Version
- 09-04-2013 | Updated with components section
- 09-05-2013 | Updated with I/O path overview section
- 09-09-2013 | Updated with converged architecture section
- 09-11-2013 | Updated with data structure section
- 09-24-2013 | Updated with data protection section
- 09-30-2013 | Updated with data locality section
- 10-01-2013 | Updated with shadow clones section
- 10-07-2013 | Updated with scalable metadata section
- 10-11-2013 | Updated with elastic dedupe engine
- 11-01-2013 | Updated with networking and I/O section
- 11-07-2013 | Updated with CVM autopathing section
- 01-23-2014 | Updated with new content structure and layout
- 02-10-2014 | Updated with storage layers and monitoring
- 02-18-2014 | Updated with array offloads sections
- 03-12-2014 | Updated with genesis
- 03-17-2014 | Updated spelling and grammar
- 03-19-2014 | Updated with apis and interfaces section
- 03-25-2014 | Updated script block formatting
- 03-26-2014 | Updated with dr & protection domain NCLI commands
- 04-15-2014 | Updated with failure domains section and 4.0 updates
- 04-23-2014 | Updated with command to echo factory_config.json
No comments:
Post a Comment