AIStore runs on a single Linux or Mac machine. Bare-metal Kubernetes and GCP/GKE Cloud-based deployments are also supported. There are numerous other options.

Generally, when deciding how to deploy a system like AIS with so many possibilities to choose from, a good place to start would be answering the following two fundamental questions:

  • what’s the dataset size, or sizes?
  • what hardware will I use?

For datasets, say, below 50TB a single host may suffice and should, therefore, be considered a viable option. On the other hand, the Cloud deployment option may sound attractive for its ubiquitous convenience and for not thinking about the hardware and the sizes - at least, not right away.

Note as well that you can always start small: a single-host deployment, a 3-node cluster in the Cloud or on-premises, etc. AIStore supports many options to inter-connect existing clusters - the capability called unified global namespace - or migrate existing datasets (on-demand or via supported storage services). For introductions and further pointers, please refer to the AIStore Overview.


AIStore runs on commodity Linux machines with no special requirements whatsoever. It is expected that within a given cluster, all AIS targets are identical, hardware-wise.

  • Linux distribution with GCC, sysstat, attr and util-linux packages(**)
  • Linux kernel 5.15+
  • Go 1.21 or later
  • Extended attributes (xattrs - see next section)
  • Optionally, Amazon (AWS), Google Cloud Platform (GCP), and/or Azure Cloud Storage accounts.

(**) Mac is also supported albeit in a limited (development only) way.

See also:


Depending on your Linux distribution, you may or may not have GCC, sysstat, and/or attr packages. These packages must be installed.

Speaking of distributions, our current default recommendation is Ubuntu Server 20.04 LTS. But Ubuntu 18.04 and CentOS 8.x (or later) will also work. As well as numerous others.

For the local filesystem, we currently recommend xfs. But again, this (default) recommendation shall not be interpreted as a limitation of any kind: other fine choices include zfs, ext4, f2fs, and more.

Since AIS itself provides n-way mirroring and erasure coding, hardware RAID would not be recommended. But can be used, and will work.

The capability called extended attributes, or xattrs, is a long-time POSIX legacy supported by all mainstream filesystems with no exceptions. Unfortunately, xattrs may not always be enabled in the Linux kernel configurations - the fact that can be easily found out by running the setfattr command.

If disabled, please make sure to enable xattrs in your Linux kernel configuration. To quickly check:

$ touch foo
$ setfattr -n -v ttt foo
$ getfattr -n foo


For developers, there’s also macOS aka Darwin option. Certain capabilities related to querrying the state and status of local hardware resources (memory, CPU, disks) may be missing. In fact, it is easy to review specifics with a quick check on the sources:

$ find . -name "*darwin*"


Benchmarking and stress-testing is also being done on Linux only - another reason to consider Linux (and only Linux) for production deployments.

The rest of this document is structured as follows:

Table of Contents

Local Playground

If you’re looking for speedy evaluation, want to experiment with supported features, get a feel of initial usage, or development - for any and all of these reasons running AIS from its GitHub source might be a good option.

Hence, we introduced and keep maintaining Local Playground - one of the several supported deployment options.

Some of the most popular deployment options are also summarized in this table. The list includes Local Playground, and its complementary guide here.

Local Playground is not intended for production and is not meant to provide optimal performance.

To run AIStore from source, one would typically need to have Go: compiler, linker, tools, and required packages. However:

CROSS_COMPILE option (see below) can be used to build AIStore without having (to install) Go and its toolchain (requires Docker).

To install Go(lang) on Linux:

  • Download the latest go1.21.<x>.linux-amd64.tar.gz from Go downloads
  • Follow installation instructions
  • Or simply run: tar -C /usr/local -xzf go1.21.<x>.linux-amd64.tar.gz and add /usr/local/go/bin to $PATH

Next, if not done yet, export the GOPATH environment variable.

Here’s an additional 5-minute introduction that talks more in-depth about setting up the Go environment variables.

Once done, we can run AIS as follows:

Step 1: Clone the AIStore repository and preload dependencies

We want to clone the repository into the following path so we can access some of the associated binaries through the environment variables we set up earlier.

$ cd $GOPATH/src/
$ git clone
$ cd aistore
# Optionally, run `make mod-tidy` to preload dependencies

Step 2: Deploy cluster and verify the running status using ais cli

NOTE: For a local deployment, we do not need production filesystem paths. For more information, read about configuration basics. If you need a physical disk or virtual block device, you must add them to the fspaths config. See running local playground with emulated disks for more information.

Many useful commands are provided via top Makefile (for details, see Make section below).

In particular, we can use make to deploy our very first 1-node cluster:

$ make kill clean cli deploy <<< $'1\n1\n1\nn\nn\nn\n0\n'

OR (same):

$ make kill clean # kill previous cluster and clean any binaries
$ make cli # build ais cli
$ make deploy
Enter number of storage targets:
Enter number of proxies (gateways):
Number of local mountpaths (enter 0 for preconfigured filesystems):
Select backend providers:
Amazon S3: (y/n) ?
Google Cloud Storage: (y/n) ?
Azure: (y/n) ?
Loopback device size, e.g. 10G, 100M (press Enter to skip):

You should see a successful deployment in the output when the AIS proxy is listening on port 8080 (unless changed in config).

Building aisnode 2ce1aaeb1 [build tags: mono]
Proxy is listening on port: 8080
Primary endpoint: http://localhost:8080

Note that with no arguments also builds AIStore binaries (such as aisnode and ais CLI). You can pass in arguments to configure the same options that the make deploy command above uses.

WARNING: Note that the --cleanup option will wipe all AIS related files and binaries. You do not need to use the --cleanup option if you just want to run the local playground with a different number of nodes or proxies. The same warning applies to running make clean as well. If you’d like to just restart the cluster without loss of data, just kill the cluster using make kill and make run to restart cluster without recreating daemon configs.

$ ./scripts/ --target-cnt 1 --proxy-cnt 1 --mountpath-cnt 1 --deployment local --cleanup

We can verify that the cluster is running using:

$ ais show cluster

# Example output below
p[FAEp8080][P]   0.13%           29.31GiB        [0.4 0.2 0.2]   -       online  3.23.rc3.2ce1aaeb1      2024-05-15T16:19:05-0700

t[hgjt8081]      0.13%           29.31GiB        39%             21.999GiB       [0.4 0.2 0.2]   -               -       online  3.23.rc3.2ce1aaeb1      2024-05-15T16:19:05-0700

   Proxies:             1
   Targets:             1 (one disk)
   Capacity:            used 14.61GiB (39%), available 22.00GiB
   Cluster Map:         version 5, UUID PrVDVYkcT, primary p[FAEp8080]
   Deployment:          dev
   Status:              2 online
   Rebalance:           n/a
   Authentication:      disabled
   Version:             3.23.rc3.2ce1aaeb1
   Build:               2024-05-15T16:19:05-0700

If you get repreated errors indicating that the cluster is taking a long time to start up, then your configuration is incorrect.

Step 3: Run aisloader tool

We can now run the aisloader tool to benchmark our new cluster.

$ make aisloader # build aisloader tool

$ aisloader -bucket=ais://abc -duration 2m -numworkers=8 -minsize=1K -maxsize=1K -pctput=100 --cleanup=false # run aisloader for 2 minutes (8 workers, 1KB size, 100% write, no cleanup)

Step 4: Run iostat (or use any of the multiple documented ways to monitor AIS performance)

$ iostat -dxm 10 sda sdb

Running Local Playground with emulated disks

Here’s a quick walk-through (with more references included below).

  • Step 1: patch deploy/dev/local/ as follows:
diff --git a/deploy/dev/local/ b/deploy/dev/local/
index c5e0e4fae..46085e19c 100755
--- a/deploy/dev/local/
+++ b/deploy/dev/local/
@@ -192,11 +192,12 @@ cat > $AIS_LOCAL_CONF_FILE <<EOL
                "port_intra_data":    "${PORT_INTRA_DATA:-10080}"
        "fspaths": {
-               $AIS_FS_PATHS
+               "/tmp/ais/mp1": "",
+               "/tmp/ais/mp2": ""
        "test_fspaths": {
                "root":     "${TEST_FSPATH_ROOT:-/tmp/ais$NEXT_TIER/}",
-               "count":    ${TEST_FSPATH_COUNT:-0},
+               "count":    0,
                "instance": ${INSTANCE:-0}
  • Step 2: deploy a single target with two loopback devices (1GB size each):
$ make kill clean cli deploy <<< $'1\n1\n2\ny\ny\nn\n1G\n'

$ mount | grep dev/loop
/dev/loop23 on /tmp/ais/mp1 type ext4 (rw,relatime)
/dev/loop25 on /tmp/ais/mp2 type ext4 (rw,relatime)
  • Step 3: observe a running cluster; notice the deployment type and the number of disks:
$ ais show cluster
p[BOxqibgv][P]   0.14%           27.28GiB        [1.2 1.1 1.1]   -       online  3.22.bf26375e5  2024-02-29T11:11:52-0500

t[IwzSpiIm]      0.14%           27.28GiB     6%           1.770GiB     [1.2 1.1 1.1]   -           -       online  3.22.bf26375e5

   Proxies:             1
   Targets:             1 (num disks: 2)
   Cluster Map:         version 4, UUID g7sPH9dTY, primary p[BOxqibgv]
   Deployment:          linux
   Status:              2 online
   Rebalance:           n/a
   Authentication:      disabled
   Version:             3.22.bf26375e5
   Build:               2024-02-29T11:11:52-0500

See also:

for developers; cluster and node configuration; supported deployments: summary table and links.

Running Local Playground remotely

AIStore (product and solution) is fully based on HTTP(S) utilizing the protocol both externally (to support both frontend interfaces and communications with remote backends) and internally, for intra-cluster streaming.

Connectivity-wise, what that means is that your local deployment at localhost:8080 can as easily run at any arbitrary HTTP(S) address.

Here’re the quick change you make to deploy Local Playground at (e.g.), whereby the main gateway’s listening port would still remain 8080 default:

diff --git a/deploy/dev/local/ b/deploy/dev/local/                                                             |
index 9198c0de4..be63f50d0 100755                                                                                                                |
--- a/deploy/dev/local/                                                                                                         |
+++ b/deploy/dev/local/                                                                                                         |
@@ -181,7 +181,7 @@ cat > $AIS_LOCAL_CONF_FILE <<EOL                                                                                             |
        "confdir": "${AIS_CONF_DIR:-/etc/ais/}",                                                                                                 |
        "log_dir":       "${AIS_LOG_DIR:-/tmp/ais$NEXT_TIER/log}",                                                                               |
        "host_net": {                                                                                                                            |
-               "hostname":                 "${HOSTNAME_LIST}",                                                                                  |
+               "hostname":                 "",                                                                                        |
                "hostname_intra_control":   "${HOSTNAME_LIST_INTRA_CONTROL}",                                                                    |
                "hostname_intra_data":      "${HOSTNAME_LIST_INTRA_DATA}",                                                                       |
                "port":               "${PORT:-8080}",                                                                                           |
diff --git a/deploy/dev/local/ b/deploy/dev/local/                                                                             |
index e0b467d82..b18361155 100755                                                                                                                |
--- a/deploy/dev/local/                                                                                                                 |
+++ b/deploy/dev/local/                                                                                                                 |
@@ -68,7 +68,7 @@ else                                                                                                                           |
   PORT_INTRA_DATA=${PORT_INTRA_DATA:-13080}                                                                                                     |
   NEXT_TIER="_next"                                                                                                                             |
 fi                                                                                                                                              |
-AIS_PRIMARY_URL="http://localhost:$PORT"                                                                                                        |
+AIS_PRIMARY_URL="$PORT"                                                                                                       |
 if $AIS_USE_HTTPS; then                                                                                                                         |
   AIS_PRIMARY_URL="https://localhost:$PORT"                                                                                                     |


AIS comes with its own build system that we use to build both standalone binaries and container images for a variety of deployment options.

The very first make command you may want to execute could as well be:

$ make help

This shows all subcommands, environment variables, and numerous usage examples, including:

# Deploy cluster locally
$ make deploy

# Stop locally deployed cluster and cleanup all cluster-related data and bucket metadata (but not cluster map)
$ make kill clean

# Stop and then deploy (non-interactively) cluster consisting of 7 targets (4 mountpaths each) and 2 proxies; build `aisnode` executable with the support for GCP and AWS backends
$ make kill deploy <<< $'7\n2\n4\ny\ny\nn\n0\n'

System environment variables

The variables include AIS_ENDPOINT, AIS_AUTHN_TOKEN_FILE, and more.

Almost in all cases, there’s an “AIS_” prefix (hint: git grep AIS_).

And in all cases with no exception, the variable takes precedence over the corresponding configuration, if exists. For instance:

AIS_ENDPOINT= ais show cluster

overrides the default endpoint as per ais config cli or (same) ais config cli --json

Endpoints are equally provided by each and every running AIS gateway (aka “proxy”) and each endpoint can be (equally) used to access the cluster. To find out what’s currently configured, run (e.g.):

$ ais config node <NODE> local host_net --json

where NODE is, effectively, any clustered proxy (that’ll show up if you type ais config node and press <TAB-TAB>).

Other variables, such as AIS_PRIMARY_EP and AIS_USE_HTTPS can prove to be useful at deployment time.

For developers, CLI ais config cluster log.modules ec xs (for instance) would allow to selectively raise and/or reduce logging verbosity on a per module bases - modules EC (erasure coding) and xactions (batch jobs) in this particular case.

To list all log modules, type ais config cluster or ais config node and press <TAB-TAB>.

Finally, there’s also HTTPS configuration (including X509 certificates and options), and the corresponding environment.

For details, please see section TLS: testing with self-signed certificates below.

Multiple deployment options

AIStore deploys anywhere anytime supporting multiple deployment options summarized and further referenced here.

All containerized deployments have their own separate Makefiles. With the exception of local playground, each specific build-able development (dev/) and production (prod/) option under the deploy folder contains a pair: {Dockerfile, Makefile}.

This separation is typically small in size and easily readable and maintainable.

Also supported is the option not to have the required Go installed and configured. To still be able to build AIS binaries without Go on your machine, make sure that you have docker and simply uncomment CROSS_COMPILE line in the top Makefile.

In the software, type of the deployment is also present in some minimal way. In particular, to overcome certain limitations of Local Playground (single disk shared by multiple targets, etc.) - we need to know the type. Which can be:

enumerated type comment
dev development
k8s Kubernetes
linux Linux

The most recently updated enumeration can be found in the source

The type shows up in the show cluster output - see example above.

Kubernetes deployments

For any Kubernetes deployments (including, of course, production deployments) please use a separate and dedicated AIS-K8s GitHub repository. The repo contains detailed Ansible playbooks that cover a variety of use cases and configurations.

Finally, the repository hosts the Kubernetes Operator project that will eventually replace Helm charts and will become the main deployment, lifecycle, and operation management “vehicle” for AIStore.

Minimal all-in-one-docker Deployment

This option has the unmatched convenience of requiring an absolute minimum time and resources - please see this README for details.

Manual deployment

You can also run make deploy in the root directory of the repository to deploy a cluster:

$ make deploy
Enter number of storage targets:
Enter number of proxies (gateways):
Number of local cache directories (enter 0 to use preconfigured filesystems):
Select backend providers:
Amazon S3: (y/n) ?
Google Cloud Storage: (y/n) ?
Azure: (y/n) ?
HDFS: (y/n) ?
Create loopback devices (note that it may take some time): (y/n) ?
Building aisnode: version=df24df77 providers=

Notice the “Cloud” prompt above and the fact that access to 3rd party Cloud storage is a deployment-time option.

Run make help for supported (make) options and usage examples, including:

# Restart a cluster of 7 targets (4 mountpaths each) and 2 proxies; utilize previously generated (pre-shutdown) local configurations
$ make restart <<< $'7\n2\n4\ny\ny\nn\n0\n'



  • make kill - terminate local AIStore.
  • make restart - shut it down and immediately restart using the existing configuration.
  • make help - show make options and usage examples.

For even more development options and tools, please refer to:

Testing your cluster

For development, health-checking a new deployment, or for any other (functional and performance testing) related reason you can run any/all of the included tests.

For example:

$ go test ./ais/test -v -run=Mirror

The go test above will create an AIS bucket, configure it as a two-way mirror, generate thousands of random objects, read them all several times, and then destroy the replicas and eventually the bucket as well.

Alternatively, if you happen to have Amazon and/or Google Cloud account, make sure to specify the corresponding (S3 or GCS) bucket name when running go test commands. For example, the following will download objects from your (presumably) S3 bucket and distribute them across AIStore:

$ BUCKET=aws://myS3bucket go test ./ais/test -v -run=download

To run all tests in the category short tests:

# using randomly named ais://nnn bucket (that will be created on the fly and destroyed in the end):
$ BUCKET=ais://nnn make test-short

# with existing Google Cloud bucket gs://myGCPbucket
$ BUCKET=gs://myGCPbucket make test-short

The command randomly shuffles existing short tests and then, depending on your platform, usually takes anywhere between 15 and 30 minutes. To terminate, press Ctrl-C at any time.

Ctrl-C or any other (kind of) abnormal termination of a running test may have a side effect of leaving some test data in the test bucket.

Kubernetes Playground

In our development and testing, we make use of Minikube and the capability, further documented here, to run the Kubernetes cluster on a single development machine. There’s a distinct advantage that AIStore extensions that require Kubernetes - such as Extract-Transform-Load, for example - can be developed rather efficiently.

Setting Up HTTPS Locally

In the end, all examples above run a bunch of local web servers that listen for plain HTTP requests. Following are quick steps for developers to engage HTTPS.

This is still a so-called local playground type deployment from scratch, whereby we are not trying to switch an existing cluster from HTTP to HTTPS, or vice versa. All we do here is deploying a brand new HTTPS-based aistore.

Note: If you need to switch an existing AIS cluster to HTTPS, please refer to these steps.

Generate Certificates

Creating a self-signed certificate along with its private key and a Certificate Authority (CA) certificate using OpenSSL involves several steps. First we create a self-signed Certificate Authority (CA) certificate (ca.crt). Then we create a Certificate Signing Request (CSR) and finally based on CSR and CA we create the server certs (server.key and server.crt).

$ openssl req -x509 -newkey rsa:2048 -keyout ca.key -out ca.crt -days 1024 -nodes -subj "/CN=localhost" -extensions v3_ca -config <(printf "[req]\ndistinguished_name=req\nx509_extensions=v3_ca\n[ v3_ca ]\nsubjectAltName=DNS:localhost,DNS:,IP:\nbasicConstraints=CA:TRUE\n")
$ openssl req -new -newkey rsa:2048 -nodes -keyout server.key -out server.csr -subj "/C=US/ST=California/L=Santa Clara/O=NVIDIA/OU=AIStore/CN=localhost" -config <(printf "[req]\ndistinguished_name=req\nreq_extensions = v3_req\n[ v3_req ]\nsubjectAltName=DNS:localhost,DNS:,IP:\n")
$ openssl x509 -req -in server.csr -CA ca.crt -CAkey ca.key -CAcreateserial -out server.crt -days 365 -sha256 -extfile <(printf "[ext]\nsubjectAltName=DNS:localhost,DNS:,IP:\nbasicConstraints=CA:FALSE\nkeyUsage=digitalSignature,nonRepudiation,keyEncipherment,dataEncipherment\nextendedKeyUsage=serverAuth,clientAuth\n") -extensions ext

Important: Ensure you specify the appropriate DNS relevant to your deployment context. For deployments intended to run locally, the domain name localhost and should be provided for the certificates to be valid.

Deploy Cluster (4 targets, 1 gateway, 6 mountpaths, AWS)

$ # shutdown prev running AIS cluster
$ make kill
$ # delete smaps of prev AIS clusters
$ find ~/.ais* -type f -name ".ais.smap" | xargs rm
$ AIS_USE_HTTPS=true AIS_SKIP_VERIFY_CRT=true AIS_SERVER_CRT=<path-to-cert>/server.crt AIS_SERVER_KEY=<path-to-key>/server.key make deploy <<< $'4\n1\n6\ny\nn\nn\n0\n'

Notice environment variables above: AIS_USE_HTTPS, AIS_SKIP_VERIFY_CRT, AIS_SERVER_CRT and AIS_SERVER_KEY.

Accessing the Cluster

To use CLI, try first any command with HTTPS-based cluster endpoint, for instance:

$ AIS_ENDPOINT= ais show cluster

But if it fails with “failed to verify certificate” message, perform a simple step to configufre CLI to skip HTTPS cert validation:

$ ais config cli set cluster.skip_verify_crt true
"cluster.skip_verify_crt" set to: "true" (was: "false")

And then try again.

TLS: testing with self-signed certificates

In the previous example, the cluster is simply ignoring SSL certificate verification due to AIS_SKIP_VERIFY_CRT being true. For a more secure setup, consider validating certificates by configuring the necessary environment variables as shown in the table below-

var name description the corresponding cluster configuration
AIS_USE_HTTPS when false, we use plain HTTP with all the TLS config (below) simply ignored “net.http.use_https”
AIS_SERVER_CRT aistore cluster X509 certificate “net.http.server_crt”
AIS_SERVER_KEY certificate’s private key “net.http.server_key”
AIS_DOMAIN_TLS NOTE: not supported, must be empty (domain, hostname, or SAN registered with the certificate) “net.http.domain_tls”
AIS_CLIENT_CA_TLS Certificate authority that authorized (signed) the certificate “net.http.client_ca_tls”
AIS_CLIENT_AUTH_TLS Client authentication during TLS handshake: a range from 0 (no authentication) to 4 (request and validate client’s certificate) “net.http.client_auth_tls”
AIS_SKIP_VERIFY_CRT when true: skip X509 cert verification (usually enabled to circumvent limitations of self-signed certs) “net.http.skip_verify”


In the following example, we run https based deployment where AIS_SKIP_VERIFY_CRT is false.

$ make kill
$ # delete smaps
$ find ~/.ais* -type f -name ".ais.smap" | xargs rm
$ # substitute varibles in below files to point to correct certificates
$ source ais/test/tls-env/server.conf
$ source ais/test/tls-env/client.conf
$ AIS_USE_HTTPS=true make deploy <<< $'6\n6\n4\ny\ny\nn\n\n'

Notice that when the cluster is first time deployed server.conf environment (above) overrides aistore cluster configuration.

Environment is ignored upon cluster restarts and upgrades.

On the other hand, ais/test/tls-env/client.conf contains environment variables to override CLI config. The correspondence between environment and config names is easy to see as well.

See also: Client-side TLS environment

Build, Make, and Development Tools

As noted, the project utilizes GNU make to build and run things both locally and remotely (e.g., when deploying AIStore via Kubernetes. As the very first step, run make help for help on:

  • building AIS node binary (called aisnode) deployable as both storage target or an ais gateway (most of the time referred to as “proxy”);
  • building CLI
  • building benchmark tools.

In particular, the make provides a growing number of developer-friendly commands to:

  • deploy the AIS cluster on your local development machine;
  • run all or selected tests;
  • instrument AIS binary with race detection, CPU and/or memory profiling, and more.

Of course, local build is intended for development only. For production, there is a separate dedicated repository noted below.

In summary:

A note on conditional linkage

AIStore build supports conditional linkage of the supported remote backends: S3, GCS, Azure.

For the most recently updated list, please see 3rd party Backend providers.

To access remote data (and store it in-cluster), AIStore utilizes the respective provider’s SDK.

For Amazon S3, that would be aws-sdk-go-v2, for Azure - azure-storage-blob-go, and so on. Each SDK can be conditionally linked into aisnode executable - the decision to link or not to link is made prior to deployment.

But not only supported remote backends are conditionally linked. Overall, the following list of commented examples presents almost all supported build tags (with maybe one minor exception):

# 1) build aisnode with no build tags, no debug
$ MODE="" make node

# 2) build aisnode with no build tags but with debug
$ MODE="debug" make node

# 3) all 3 cloud backends, no debug
$ AIS_BACKEND_PROVIDERS="aws azure gcp" MODE="" make node

# 4) cloud backends, with debug
$ AIS_BACKEND_PROVIDERS="aws azure gcp" MODE="debug" make node

# 5) cloud backends, debug, statsd
## Note: if `statsd` build tag is not specified `aisnode` will get built with Prometheus support.
## For more information (including the binary choice between StatsD and Prometheus), please see docs/ and docs/
$ TAGS="aws azure gcp statsd debug" make node

# 6) statsd, debug, nethttp (note that fasthttp is used by default)
$ TAGS="nethttp statsd debug" make node

In addition, to build AuthN, CLI, and/or aisloader, run:

  • make authn
  • make cli
  • make aisloader

respectively. With each of these makes, you can also use MODE=debug - debug mode is universally supported.

Containerized Deployments: Host Resource Sharing

The following applies to all containerized deployments:

  1. AIS nodes always automatically detect containerization.
  2. If deployed as a container, each AIS node independently discovers whether its own container’s memory and/or CPU resources are restricted.
  3. Finally, the node then abides by those restrictions.

To that end, each AIS node at startup loads and parses cgroup settings for the container and, if the number of CPUs is restricted, adjusts the number of allocated system threads for its goroutines.

This adjustment is accomplished via the Go runtime GOMAXPROCS variable. For in-depth information on CPU bandwidth control and scheduling in a multi-container environment, please refer to the CFS Bandwidth Control document.

Further, given the container’s cgroup/memory limitation, each AIS node adjusts the amount of memory available for itself.

Memory limits may affect dSort performance forcing it to “spill” the content associated with in-progress resharding into local drives. The same is true for erasure-coding which also requires memory to rebuild objects from slices, etc.

For technical details on AIS memory management, please see this readme.

Assorted command lines

AIStore targets may execute (and parse the output of) the following command:

$ lsblk -Jt


In fact, prior to deploying AIS cluster on a given Linux distribution the very first time, it’d make sense to maybe run the 3 commands and check output for “invalid option” or lack of thereof.