- Administration >
- Production Notes
On this page
This page details system configurations that affect MongoDB, especially when running in production.
MongoDB Cloud Manager, a hosted service, and Ops Manager, an on-premise solution, provide monitoring, backup, and automation of MongoDB instances. See the MongoDB Cloud Manager documentation and Ops Manager documentation for more information.
MongoDB provides builds for the following supported platforms. For running in production, refer to the Recommended Platforms for operating system recommendations.
|Windows Vista/Server 2008R2/2012+||✓||✓||✓||✓||✓|
|OS X 10.7+||✓||✓||✓||✓|
Changed in version 3.2: MongoDB can now use the WiredTiger storage engine on all supported platforms.
While MongoDB supports a variety of platforms, the following operating systems are recommended for production use:
- Amazon Linux
- Debian 7.1
- RHEL / CentOS 6.2+
- SLES 11+
- Ubuntu LTS 12.04
- Ubuntu LTS 14.04
- Windows Server 2012 & 2012 R2
Use the Latest Stable Packages¶
Be sure you have the latest stable release.
Use 64-bit Builds¶
Always use 64-bit builds for production.
Starting in MongoDB 3.2, 32-bit binaries are deprecated and will be unavailable in future releases.
Although the 32-bit builds exist for Linux and Windows, they are unsuitable for production deployments. 32-bit builds also do not support the WiredTiger storage engine. For more information, see the 32-bit limitations page
The files in the dbPath directory must correspond to the configured storage engine. mongod will not start if dbPath contains data files created by a storage engine other than the one specified by --storageEngine.
Changed in version 3.2: As of MongoDB 3.2, MongoDB uses the WiredTiger storage engine by default.
Changed in version 3.0: Beginning with MongoDB 3.0, MMAPv1 provides collection-level locking: All collections have a unique readers-writer lock that allows multiple clients to modify documents in different collections at the same time.
For MongoDB versions 2.2 through 2.6 series, each database has a readers-writer lock that allows concurrent read access to a database, but gives exclusive access to a single write operation per database. See the Concurrency page for more information. In earlier versions of MongoDB, all write operations contended for a single readers-writer lock for the entire mongod instance.
WiredTiger supports concurrent access by readers and writers to the documents in a collection. Clients can read documents while write operations are in progress, and multiple threads can modify different documents in a collection at the same time.
Allocate Sufficient RAM and CPU provides information about how WiredTiger takes advantage of multiple CPU cores and how to improve operation throughput.
MongoDB uses write ahead logging to an on-disk journal. Journaling guarantees that MongoDB can quickly recover write operations that were written to the journal but not written to data files in cases where mongod terminated due to a crash or other serious failure.
New in version 3.2.
To use a read concern level of "majority", you must use the WiredTiger storage engine and start the mongod instances with the --enableMajorityReadConcern command line option (or the replication.enableMajorityReadConcern setting if using a configuration file).
Write concern describes the level of acknowledgement requested from MongoDB for write operations. The level of the write concerns affects how quickly the write operation returns. When write operations have a weak write concern, they return quickly. With stronger write concerns, clients must wait after sending a write operation until MongoDB confirms the write operation at the requested write concern level. With insufficient write concerns, write operations may appear to a client to have succeeded, but may not persist in some cases of server failure.
See the Write Concern document for more information about choosing an appropriate write concern level for your deployment.
Use Trusted Networking Environments¶
Always run MongoDB in a trusted environment, with network rules that prevent access from all unknown machines, systems, and networks. As with any sensitive system that is dependent on network access, your MongoDB deployment should only be accessible to specific systems that require access, such as application servers, monitoring services, and other MongoDB components.
By default, authorization is not enabled, and mongod assumes a trusted environment. Enable authorization mode as needed. For more information on authentication mechanisms supported in MongoDB as well as authorization in MongoDB, see Authentication and Role-Based Access Control.
For additional information and considerations on security, refer to the documents in the Security Section, specifically:
For Windows users, consider the Windows Server Technet Article on TCP Configuration when deploying MongoDB on Windows.
Disable HTTP Interface¶
MongoDB provides an HTTP interface to check the status of the server and, optionally, run queries. The HTTP interface is disabled by default. Do not enable the HTTP interface in production environments.
Deprecated since version 3.2: HTTP interface for MongoDB
Manage Connection Pool Sizes¶
Avoid overloading the connection resources of a mongod or mongos instance by adjusting the connection pool size to suit your use case. Start at 110-115% of the typical number of current database requests, and modify the connection pool size as needed. Refer to the Connection Pool Options for adjusting the connection pool size.
See also Allocate Sufficient RAM and CPU.
MongoDB is designed specifically with commodity hardware in mind and has few hardware requirements or limitations. MongoDB’s core components run on little-endian hardware, primarily x86/x86_64 processors. Client libraries (i.e. drivers) can run on big or little endian systems.
Allocate Sufficient RAM and CPU¶
Due to its concurrency model, the MMAPv1 storage engine does not require many CPU cores. As such, increasing the number of cores can improve performance but does not provide significant return.
Increasing the amount of RAM accessible to MongoDB may help reduce the frequency of page faults.
The WiredTiger storage engine is multithreaded and can take advantage of additional CPU cores. Specifically, the total number of active threads (i.e. concurrent operations) relative to the number of available CPUs can impact performance:
- Throughput increases as the number of concurrent active operations increases up to the number of CPUs.
- Throughput decreases as the number of concurrent active operations exceeds the number of CPUs by some threshold amount.
The threshold depends on your application. You can determine the optimum number of concurrent active operations for your application by experimenting and measuring throughput. The output from mongostat provides statistics on the number of active reads/writes in the (ar|aw) column.
With WiredTiger, MongoDB utilizes both the WiredTiger internal cache and the filesystem cache.
Changed in version 3.2: Starting in MongoDB 3.2, the WiredTiger internal cache, by default, will use the larger of either:
- 60% of RAM minus 1 GB, or
- 1 GB.
For systems with up to 10 GB of RAM, the new default setting is less than or equal to the 3.0 default setting (For MongoDB 3.0, the WiredTiger internal cache uses either 1 GB or half of the installed physical RAM, whichever is larger).
For systems with more than 10 GB of RAM, the new default setting is greater than the 3.0 setting.
Via the filesystem cache, MongoDB automatically uses all free memory that is not used by the WiredTiger cache or by other processes. Data in the filesystem cache is compressed.
To adjust the size of the WiredTiger internal cache, see storage.wiredTiger.engineConfig.cacheSizeGB and --wiredTigerCacheSizeGB. Avoid increasing the WiredTiger internal cache size above its default value.
The storage.wiredTiger.engineConfig.cacheSizeGB limits the size of the WiredTiger internal cache. The operating system will use the available free memory for filesystem cache, which allows the compressed MongoDB data files to stay in memory. In addition, the operating system will use any free RAM to buffer file system blocks and file system cache.
To accommodate the additional consumers of RAM, you may have to decrease WiredTiger internal cache size.
The default WiredTiger internal cache size value assumes that there is a single mongod instance per machine. If a single machine contains multiple MongoDB instances, then you should decrease the setting to accommodate the other mongod instances.
If you run mongod in a container (e.g. lxc, cgroups, Docker, etc.) that does not have access to all of the RAM available in a system, you must set storage.wiredTiger.engineConfig.cacheSizeGB to a value less than the amount of RAM available in the container. The exact amount depends on the other processes running in the container.
Use Solid State Disks (SSDs)¶
MongoDB has good results and a good price-performance ratio with SATA SSD (Solid State Disk).
Use SSD if available and economical. Spinning disks can be performant, but SSDs’ capacity for random I/O operations works well with the update model of MMAPv1.
Commodity (SATA) spinning drives are often a good option, as the random I/O performance increase with more expensive spinning drives is not that dramatic (only on the order of 2x). Using SSDs or increasing RAM may be more effective in increasing I/O throughput.
MongoDB and NUMA Hardware¶
Running MongoDB on a system with Non-Uniform Access Memory (NUMA) can cause a number of operational problems, including slow performance for periods of time and high system process usage.
When running MongoDB servers and clients on NUMA hardware, you should configure a memory interleave policy so that the host behaves in a non-NUMA fashion. MongoDB checks NUMA settings on start up when deployed on Linux (since version 2.0) and Windows (since version 2.6) machines. If the NUMA configuration may degrade performance, MongoDB prints a warning.
- The MySQL “swap insanity” problem and the effects of NUMA post, which describes the effects of NUMA on databases. The post introduces NUMA and its goals, and illustrates how these goals are not compatible with production databases. Although the blog post addresses the impact of NUMA for MySQL, the issues for MongoDB are similar.
- NUMA: An Overview.
Configuring NUMA on Windows¶
On Windows, memory interleaving must be enabled through the machine’s BIOS. Consult your system documentation for details.
Configuring NUMA on Linux¶
When running MongoDB on Linux, you should disable zone reclaim in the sysctl settings using one of the following commands:
echo 0 | sudo tee /proc/sys/vm/zone_reclaim_mode
sudo sysctl -w vm.zone_reclaim_mode=0
Then, you should use numactl to start your mongod instances, including the config servers, mongos instances, and any clients. If you do not have the numactl command, refer to the documentation for your operating system to install the numactl package.
The following operation demonstrates how to start a MongoDB instance using numactl:
numactl --interleave=all <path> <options>
The <path> is the path to the program you are starting and the <options> are any optional arguments to pass to the program.
To fully disable NUMA behavior, you must perform both operations. For more information, see the Documentation for /proc/sys/vm/*.
Disk and Storage Systems¶
Assign swap space for your systems. Allocating swap space can avoid issues with memory contention and can prevent the OOM Killer on Linux systems from killing mongod.
For the MMAPv1 storage engine, the method mongod uses to map files to memory ensures that the operating system will never store MongoDB data in swap space. On Windows systems, using MMAPv1 requires extra swap space due to commitment limits. For details, see MongoDB on Windows.
For the WiredTiger storage engine, given sufficient memory pressure, WiredTiger may store data in swap space.
Most MongoDB deployments should use disks backed by RAID-10.
RAID-5 and RAID-6 do not typically provide sufficient performance to support a MongoDB deployment.
Avoid RAID-0 with MongoDB deployments. While RAID-0 provides good write performance, it also provides limited availability and can lead to reduced performance on read operations, particularly when using Amazon’s EBS volumes.
With the MMAPv1 storage engine, the Network File System protocol (NFS) is not recommended as you may see performance problems when both the data files and the journal files are hosted on NFS. You may experience better performance if you place the journal on local or iscsi volumes.
With the WiredTiger storage engine, WiredTiger objects may be stored on remote file systems if the remote file system conforms to ISO/IEC 9945-1:1996 (POSIX.1). Because remote file systems are often slower than local file systems, using a remote file system for storage may degrade performance.
If you decide to use NFS, add the following NFS options to your /etc/fstab file: bg, nolock, and noatime.
Separate Components onto Different Storage Devices¶
For improved performance, consider separating your database’s data, journal, and logs onto different storage devices, based on your application’s access and write pattern. Mount the components as separate filesystems and use symbolic links to map each component’s path to the device storing it.
For the WiredTiger storage engine, you can also store the indexes on a different storage device. See storage.wiredTiger.engineConfig.directoryForIndexes.
Using different storage devices will affect your ability to create snapshot-style backups of your data, since the files will be on different devices and volumes.
Scheduling for Virtual or Cloud Hosted Devices¶
For local block devices attached to a virtual machine instance via the hypervisor or hosted by a cloud hosting provider, the guest operating system should use a noop scheduler for best performance. The noop scheduler allows the operating system to defer I/O scheduling to the underlying hypervisor.
Scheduling for Physical Servers¶
For physical servers, the operating system should use a deadline scheduler. The deadline scheduler caps maximum latency per request and maintains a good disk throughput that is best for disk-intensive database applications.
WiredTiger can compress collection data using either snappy or zlib compression library. snappy provides a lower compression rate but has little performance cost, whereas zlib provides better compression rate but has a higher performance cost.
By default, WiredTiger uses snappy compression library. To change the compression setting, see storage.wiredTiger.collectionConfig.blockCompressor.
WiredTiger uses prefix compression on all indexes by default.
Platform Specific Considerations¶
MongoDB uses the GNU C Library (glibc) if available on a system. MongoDB requires version at least glibc-2.12-1.2.el6 to avoid a known bug with earlier versions. For best results use at least version 2.13.
MongoDB on Linux¶
Kernel and File Systems¶
When running MongoDB in production on Linux, you should use Linux kernel version 2.6.36 or later, with either the XFS or EXT4 filesystem. If possible, use XFS as it generally performs better with MongoDB.
With the WiredTiger storage engine, use of XFS is strongly recommended to avoid performance issues that may occur when using EXT4 with WiredTiger.
With the MMAPv1 storage engine, MongoDB preallocates its database files before using them and often creates large files. As such, you should use the XFS or EXT4 file systems. If possible, use XFS as it generally performs better with MongoDB.
- In general, if you use the XFS file system, use at least version 2.6.25 of the Linux Kernel.
- If you use the EXT4 file system, use at least version 2.6.28 of the Linux Kernel.
- On Red Hat Enterprise Linux and CentOS, use at least version 2.6.18-194 of the Linux kernel.
fsync() on Directories¶
MongoDB requires a filesystem that supports fsync() on directories. For example, HGFS and Virtual Box’s shared folders do not support this operation.
For all MongoDB deployments:
- Use the Network Time Protocol (NTP) to synchronize time among your hosts. This is especially important in sharded clusters.
For the WiredTiger and MMAPv1 storage engines, consider the following recommendations:
Turn off atime for the storage volume containing the database files.
Set the file descriptor limit, -n, and the user process limit (ulimit), -u, above 20,000, according to the suggestions in the ulimit reference. A low ulimit will affect MongoDB when under heavy use and can produce errors and lead to failed connections to MongoDB processes and loss of service.
Disable Transparent Huge Pages. MongoDB performs better with normal (4096 bytes) virtual memory pages. See Transparent Huge Pages Settings.
Disable NUMA in your BIOS. If that is not possible, see MongoDB on NUMA Hardware.
Problems have been reported when using MongoDB with SELinux enabled. To avoid issues, disable SELinux when possible.
If you are using SELinux on Red Hat, you must configure SELinux to be able to run MongoDB. See: Configure SELinux for MongoDB and Configure SELinux for MongoDB Enterprise for the required configuration.
For the WiredTiger storage engine:
- Set the readahead setting to 0 or 16. Setting a higher readahead benefits sequential I/O operations. However, since MongoDB disk access patterns are generally random, setting a higher readahead provides limited benefit. As such, for most workloads, a readahead of 0 or 16 provides optimal MongoDB performance.
For the MMAPv1 storage engine:
Ensure that readahead settings for the block devices that store the database files are appropriate. For random access use patterns, set low readahead values. A readahead of 32 (16 kB) often works well.
For a standard block device, you can run sudo blockdev --report to get the readahead settings and sudo blockdev --setra <value> <device> to change the readahead settings. Refer to your specific operating system manual for more information.
MongoDB and TLS/SSL Libraries¶
On Linux platforms, you may observe one of the following statements in the MongoDB log:
<path to SSL libs>/libssl.so.<version>: no version information available (required by /usr/bin/mongod) <path to SSL libs>/libcrypto.so.<version>: no version information available (required by /usr/bin/mongod)
These warnings indicate that the system’s TLS/SSL libraries are different from the TLS/SSL libraries that the mongod was compiled against. Typically these messages do not require intervention; however, you can use the following operations to determine the symbol versions that mongod expects:
objdump -T <path to mongod>/mongod | grep " SSL_" objdump -T <path to mongod>/mongod | grep " CRYPTO_"
These operations will return output that resembles one the of the following lines:
0000000000000000 DF *UND* 0000000000000000 libssl.so.10 SSL_write 0000000000000000 DF *UND* 0000000000000000 OPENSSL_1.0.0 SSL_write
The last two strings in this output are the symbol version and symbol name. Compare these values with the values returned by the following operations to detect symbol version mismatches:
objdump -T <path to TLS/SSL libs>/libssl.so.1* objdump -T <path to TLS/SSL libs>/libcrypto.so.1*
This procedure is neither exact nor exhaustive: many symbols used by mongod from the libcrypto library do not begin with CRYPTO_.
MongoDB on Windows¶
MongoDB 3.0 Using WiredTiger¶
For MongoDB instances using the WiredTiger storage engine, performance on Windows is comparable to performance on Linux.
MongoDB Using MMAPv1¶
Install Hotfix for MongoDB 2.6.6 and Later¶
Microsoft has released a hotfix for Windows 7 and Windows Server 2008 R2, KB2731284, that repairs a bug in these operating systems’ use of memory-mapped files that adversely affects the performance of MongoDB using the MMAPv1 storage engine.
Install this hotfix to obtain significant performance improvements on MongoDB 2.6.6 and later releases in the 2.6 series, which use MMAPv1 exclusively, and on 3.0 and later when using MMAPv1 as the storage engine.
Configure Windows Page File For MMAPv1¶
Configure the page file such that the minimum and maximum page file size are equal and at least 32 GB. Use a multiple of this size if, during peak usage, you expect concurrent writes to many databases or collections. However, the page file size does not need to exceed the maximum size of the database.
A large page file is needed as Windows requires enough space to accommodate all regions of memory mapped files made writable during peak usage, regardless of whether writes actually occur.
The page file is not used for database storage and will not receive writes during normal MongoDB operation. As such, the page file will not affect performance, but it must exist and be large enough to accommodate Windows’ commitment rules during peak database use.
Dynamic page file sizing is too slow to accommodate the rapidly fluctuating commit charge of an active MongoDB deployment. This can result in transient overcommitment situations that may lead to abrupt server shutdown with a VirtualProtect error 1455.
MongoDB on Virtual Environments¶
This section describes considerations when running MongoDB in some of the more common virtual environments.
For all platforms, consider Scheduling.
MongoDB is compatible with EC2. MongoDB Cloud Manager provides integration with Amazon Web Services (AWS) and lets you deploy new EC2 instances directly from MongoDB Cloud Manager. See Configure AWS Integration for more details.
Use Premium Storage. Microsoft Azure offers two general types of storage: Standard storage, and Premium storage. MongoDB on Azure has better performance when using Premium storage than it does with Standard storage.
For all MMAPv1 MongoDB deployments using Azure, you must mount the volume that hosts the mongod instance’s dbPath with the Host Cache Preference READ/WRITE. This applies to all Azure deployments running MMAPv1, using any guest operating system.
If your volumes have inappropriate cache settings, MongoDB may eventually shut down with the following error:
[DataFileSync] FlushViewOfFile for <data file> failed with error 1 ... [DataFileSync] Fatal Assertion 16387
The performance characteristics of MongoDB may change with READ/WRITE caching enabled.
The TCP keepalive on the Azure load balancer is 240 seconds by default, which can cause it to silently drop connections if the TCP keepalive on your Azure systems is greater than this value. You should set tcp_keepalive_time to 120 to ameliorate this problem.
On Linux systems:
To view the keep alive setting, you can use one of the following commands:
The value is measured in seconds.
To change the tcp_keepalive_time value, you can use one of the following command:
sudo sysctl -w net.ipv4.tcp_keepalive_time=<value>
echo <value> | sudo tee /proc/sys/net/ipv4/tcp_keepalive_time
These operations do not persist across system reboots. To persist the setting, add the following line to /etc/sysctl.conf:
net.ipv4.tcp_keepalive_time = <value>
For Windows systems:
To view the keep alive setting, issue the following command:
reg query HKLM\SYSTEM\CurrentControlSet\Services\Tcpip\Parameters /v KeepAliveTime
The registry value is not present by default. The system default, used if the value is absent, is 7200000 milliseconds or 0x6ddd00 in hexadecimal.
To change the KeepAliveTime value, use the following command in an Administrator Command Prompt, where <value> is expressed in hexadecimal (e.g. 0x0124c0 is 120000):
reg add HKLM\SYSTEM\CurrentControlSet\Services\Tcpip\Parameters\ /v KeepAliveTime /d <value>
Windows users should consider the Windows Server Technet Article on KeepAliveTime for more information on setting keep alive for MongoDB deployments on Windows systems.
MongoDB is compatible with VMWare.
VMWare supports memory overcommitment, where you can assign more memory to your virtual machines than the physical machine has available. When memory is overcommitted, the hypervisor reallocates memory between the virtual machines. VMWare’s balloon driver (vmmemctl) reclaims the pages that are considered least valuable. The balloon driver resides inside the guest operating system. When the balloon driver expands, it may induce the guest operating system to reclaim memory from guest applications, which can interfere with MongoDB’s memory management and affect MongoDB’s performance.
You can disable the balloon driver and VMWare’s memory overcommitment feature to mitigate these problems. However, disabling the balloon driver can cause the hypervisor to use its swap, as there is no other available mechanism to perform the memory reclamation. Accessing data in swap is much slower than accessing data in memory, which can in turn affect performance. Instead of disabling the balloon driver and memory overcommitment features, map and reserve the full amount of memory for the virtual machine running MongoDB. This ensures that the balloon will not be inflated in the local operating system if there is memory pressure in the hypervisor due to an overcommitted configuration.
When using MongoDB with VMWare, ensure that the CPU reservation does not exceed more than 2 virtual CPUs per physical core.
Disable VMWare’s Migration with vMotion (“live migration”). The live migration of a virtual machine can cause performance problems and affect replica set and sharded cluster high availability mechanisms.
It is possible to clone a virtual machine running MongoDB. You might use this function to spin up a new virtual host to add as a member of a replica set. If you clone a VM with journaling enabled, the clone snapshot will be valid. If not using journaling, first stop mongod, then clone the VM, and finally, restart mongod.
MongoDB is compatible with KVM.
KVM supports memory overcommitment, where you can assign more memory to your virtual machines than the physical machine has available. When memory is overcommitted, the hypervisor reallocates memory between the virtual machines. KVM’s balloon driver reclaims the pages that are considered least valuable. The balloon driver resides inside the guest operating system. When the balloon driver expands, it may induce the guest operating system to reclaim memory from guest applications, which can interfere with MongoDB’s memory management and affect MongoDB’s performance.
You can disable the balloon driver and KVM’s memory overcommitment feature to mitigate these problems. However, disabling the balloon driver can cause the hypervisor to use its swap, as there is no other available mechanism to perform the memory reclamation. Accessing data in swap is much slower than accessing data in memory, which can in turn affect performance. Instead of disabling the balloon driver and memory overcommitment features, map and reserve the full amount of memory for the virtual machine running MongoDB. This ensures that the balloon will not be inflated in the local operating system if there is memory pressure in the hypervisor due to an overcommitted configuration.
When using MongoDB with KVM, ensure that the CPU reservation does not exceed more than 2 virtual CPUs per physical core.
On Linux, use the iostat command to check if disk I/O is a bottleneck for your database. Specify a number of seconds when running iostat to avoid displaying stats covering the time since server boot.
For example, the following command will display extended statistics and the time for each displayed report, with traffic in MB/s, at one second intervals:
iostat -xmt 1
Key fields from iostat:
- %util: this is the most useful field for a quick check, it indicates what percent of the time the device/drive is in use.
- avgrq-sz: average request size. Smaller number for this value reflect more random IO operations.