2019-03-31 18:32:15 +00:00
.. _hardware-recommendations:
2012-05-03 03:31:35 +00:00
==========================
Hardware Recommendations
==========================
2012-07-10 15:14:42 +00:00
2012-09-07 03:31:46 +00:00
Ceph was designed to run on commodity hardware, which makes building and
maintaining petabyte-scale data clusters economically feasible.
When planning out your cluster hardware, you will need to balance a number
of considerations, including failure domains and potential performance
issues. Hardware planning should include distributing Ceph daemons and
other processes that use Ceph across many hosts. Generally, we recommend
running Ceph daemons of a specific type on a host configured for that type
of daemon. We recommend using other hosts for processes that utilize your
2013-10-23 00:11:07 +00:00
data cluster (e.g., OpenStack, CloudStack, etc).
2012-04-11 18:21:43 +00:00
2018-07-17 10:13:58 +00:00
.. tip :: Check out the `Ceph blog`_ too.
2013-02-01 00:34:02 +00:00
2012-09-17 23:25:54 +00:00
CPU
===
Ceph metadata servers dynamically redistribute their load, which is CPU
intensive. So your metadata servers should have significant processing power
2014-05-02 00:04:43 +00:00
(e.g., quad core or better CPUs). Ceph OSDs run the :term: `RADOS` service, calculate
data placement with :term: `CRUSH` , replicate data, and maintain their own copy of the
2012-09-17 23:25:54 +00:00
cluster map. Therefore, OSDs should have a reasonable amount of processing power
2013-02-01 00:34:02 +00:00
(e.g., dual core processors). Monitors simply maintain a master copy of the
2012-09-17 23:25:54 +00:00
cluster map, so they are not CPU intensive. You must also consider whether the
host machine will run CPU-intensive processes in addition to Ceph daemons. For
example, if your hosts will run computing VMs (e.g., OpenStack Nova), you will
need to ensure that these other processes leave sufficient processing power for
Ceph daemons. We recommend running additional CPU-intensive processes on
separate hosts.
RAM
===
2018-10-05 22:21:53 +00:00
Generally, more RAM is better.
Monitors and managers (ceph-mon and ceph-mgr)
---------------------------------------------
Monitor and manager daemon memory usage generally scales with the size of the
cluster. For small clusters, 1-2 GB is generally sufficient. For
large clusters, you should provide more (5-10 GB). You may also want
to consider tuning settings like `` mon_osd_cache_size `` or
`` rocksdb_cache_size `` .
Metadata servers (ceph-mds)
---------------------------
The metadata daemon memory utilization depends on how much memory its cache is
configured to consume. We recommend 1 GB as a minimum for most systems. See
`` mds_cache_memory `` .
OSDs (ceph-osd)
---------------
By default, OSDs that use the BlueStore backend require 3-5 GB of RAM. You can
adjust the amount of memory the OSD consumes with the `` osd_memory_target `` configuration option when BlueStore is in use. When using the legacy FileStore backend, the operating system page cache is used for caching data, so no tuning is normally needed, and the OSD memory consumption is generally related to the number of PGs per daemon in the system.
2012-09-17 23:25:54 +00:00
Data Storage
============
2013-02-01 00:34:02 +00:00
Plan your data storage configuration carefully. There are significant cost and
performance tradeoffs to consider when planning for data storage. Simultaneous
OS operations, and simultaneous request for read and write operations from
2017-06-30 17:54:18 +00:00
multiple daemons against a single drive can slow performance considerably.
2013-02-01 00:34:02 +00:00
.. important :: Since Ceph has to write all data to the journal before it can
2016-04-12 12:49:55 +00:00
send an ACK (for XFS at least), having the journal and OSD
2013-02-01 00:34:02 +00:00
performance in balance is really important!
Hard Disk Drives
----------------
OSDs should have plenty of hard disk drive space for object data. We recommend a
minimum hard disk drive size of 1 terabyte. Consider the cost-per-gigabyte
advantage of larger disks. We recommend dividing the price of the hard disk
drive by the number of gigabytes to arrive at a cost per gigabyte, because
larger drives may have a significant impact on the cost-per-gigabyte. For
example, a 1 terabyte hard disk priced at $75.00 has a cost of $0.07 per
gigabyte (i.e., $75 / 1024 = 0.0732). By contrast, a 3 terabyte hard disk priced
at $150.00 has a cost of $0.05 per gigabyte (i.e., $150 / 3072 = 0.0488). In the
foregoing example, using the 1 terabyte disks would generally increase the cost
2013-10-23 00:39:35 +00:00
per gigabyte by 40%--rendering your cluster substantially less cost efficient.
Also, the larger the storage drive capacity, the more memory per Ceph OSD Daemon
you will need, especially during rebalancing, backfilling and recovery. A
general rule of thumb is ~1GB of RAM for 1TB of storage space.
2013-02-01 00:34:02 +00:00
.. tip :: Running multiple OSDs on a single disk--irrespective of partitions--is
**NOT** a good idea.
.. tip :: Running an OSD and a monitor or a metadata server on a single
disk--irrespective of partitions--is **NOT** a good idea either.
Storage drives are subject to limitations on seek time, access time, read and
write times, as well as total throughput. These physical limitations affect
overall system performance--especially during recovery. We recommend using a
dedicated drive for the operating system and software, and one drive for each
2013-10-23 00:39:35 +00:00
Ceph OSD Daemon you run on the host. Most "slow OSD" issues arise due to running
an operating system, multiple OSDs, and/or multiple journals on the same drive.
2013-02-01 00:34:02 +00:00
Since the cost of troubleshooting performance issues on a small cluster likely
exceeds the cost of the extra disk drives, you can accelerate your cluster
design planning by avoiding the temptation to overtax the OSD storage drives.
2013-10-23 00:39:35 +00:00
You may run multiple Ceph OSD Daemons per hard disk drive, but this will likely
lead to resource contention and diminish the overall throughput. You may store a
journal and object data on the same drive, but this may increase the time it
takes to journal a write and ACK to the client. Ceph must write to the journal
2017-06-30 17:54:18 +00:00
before it can ACK the write.
2013-02-01 00:34:02 +00:00
Ceph best practices dictate that you should run operating systems, OSD data and
OSD journals on separate drives.
Solid State Drives
------------------
One opportunity for performance improvement is to use solid-state drives (SSDs)
to reduce random access time and read latency while accelerating throughput.
2013-02-01 19:30:30 +00:00
SSDs often cost more than 10x as much per gigabyte when compared to a hard disk
drive, but SSDs often exhibit access times that are at least 100x faster than a
hard disk drive.
2013-02-01 00:34:02 +00:00
2017-07-27 16:59:00 +00:00
SSDs do not have moving mechanical parts so they are not necessarily subject to
2013-02-01 00:34:02 +00:00
the same types of limitations as hard disk drives. SSDs do have significant
limitations though. When evaluating SSDs, it is important to consider the
performance of sequential reads and writes. An SSD that has 400MB/s sequential
write throughput may have much better performance than an SSD with 120MB/s of
sequential write throughput when storing multiple journals for multiple OSDs.
2013-01-18 06:13:12 +00:00
.. important :: We recommend exploring the use of SSDs to improve performance.
However, before making a significant investment in SSDs, we **strongly
recommend** both reviewing the performance metrics of an SSD and testing the
2013-02-01 00:34:02 +00:00
SSD in a test configuration to gauge performance.
Since SSDs have no moving mechanical parts, it makes sense to use them in the
2013-10-23 00:39:35 +00:00
areas of Ceph that do not use a lot of storage space (e.g., journals).
Relatively inexpensive SSDs may appeal to your sense of economy. Use caution.
Acceptable IOPS are not enough when selecting an SSD for use with Ceph. There
are a few important performance considerations for journals and SSDs:
2013-02-01 00:34:02 +00:00
- **Write-intensive semantics:** Journaling involves write-intensive semantics,
so you should ensure that the SSD you choose to deploy will perform equal to
or better than a hard disk drive when writing data. Inexpensive SSDs may
introduce write latency even as they accelerate access time, because
2013-02-01 19:30:30 +00:00
sometimes high performance hard drives can write as fast or faster than
some of the more economical SSDs available on the market!
2013-02-01 00:34:02 +00:00
2013-02-01 19:30:30 +00:00
- **Sequential Writes:** When you store multiple journals on an SSD you must
2013-02-01 00:34:02 +00:00
consider the sequential write limitations of the SSD too, since they may be
2013-02-01 19:30:30 +00:00
handling requests to write to multiple OSD journals simultaneously.
2013-02-01 00:34:02 +00:00
- **Partition Alignment:** A common problem with SSD performance is that
2013-02-01 19:30:30 +00:00
people like to partition drives as a best practice, but they often overlook
proper partition alignment with SSDs, which can cause SSDs to transfer data
much more slowly. Ensure that SSD partitions are properly aligned.
2013-02-01 00:34:02 +00:00
While SSDs are cost prohibitive for object storage, OSDs may see a significant
2013-02-01 19:30:30 +00:00
performance improvement by storing an OSD's journal on an SSD and the OSD's
object data on a separate hard disk drive. The `` osd journal `` configuration
setting defaults to `` /var/lib/ceph/osd/$cluster-$id/journal `` . You can mount
this path to an SSD or to an SSD partition so that it is not merely a file on
the same disk as the object data.
2013-02-01 00:34:02 +00:00
2013-02-01 19:30:30 +00:00
One way Ceph accelerates CephFS filesystem performance is to segregate the
storage of CephFS metadata from the storage of the CephFS file contents. Ceph
provides a default `` metadata `` pool for CephFS metadata. You will never have to
create a pool for CephFS metadata, but you can create a CRUSH map hierarchy for
your CephFS metadata pool that points only to a host's SSD storage media. See
`Mapping Pools to Different Types of OSDs`_ for details.
2013-02-01 00:34:02 +00:00
Controllers
-----------
Disk controllers also have a significant impact on write throughput. Carefully,
consider your selection of disk controllers to ensure that they do not create
a performance bottleneck.
2018-07-17 10:13:58 +00:00
.. tip :: The `Ceph blog`_ is often an excellent source of information on Ceph
2013-02-01 00:34:02 +00:00
performance issues. See `Ceph Write Throughput 1`_ and `Ceph Write
Throughput 2`_ for additional details.
Additional Considerations
-------------------------
2013-01-18 06:13:12 +00:00
You may run multiple OSDs per host, but you should ensure that the sum of the
total throughput of your OSD hard disks doesn't exceed the network bandwidth
required to service a client's need to read or write data. You should also
2013-02-01 00:34:02 +00:00
consider what percentage of the overall data the cluster stores on each host. If
the percentage on a particular host is large and the host fails, it can lead to
problems such as exceeding the `` full ratio `` , which causes Ceph to halt
operations as a safety precaution that prevents data loss.
When you run multiple OSDs per host, you also need to ensure that the kernel
is up to date. See `OS Recommendations`_ for notes on `` glibc `` and
`` syncfs(2) `` to ensure that your hardware performs as expected when running
multiple OSDs per host.
2013-01-18 06:13:12 +00:00
2012-09-17 23:25:54 +00:00
Networks
========
We recommend that each host have at least two 1Gbps network interface
controllers (NICs). Since most commodity hard disk drives have a throughput of
2012-09-27 23:50:16 +00:00
approximately 100MB/second, your NICs should be able to handle the traffic for
2012-09-17 23:25:54 +00:00
the OSD disks on your host. We recommend a minimum of two NICs to account for a
public (front-side) network and a cluster (back-side) network. A cluster network
(preferably not connected to the internet) handles the additional load for data
replication and helps stop denial of service attacks that prevent the cluster
from achieving `` active + clean `` states for placement groups as OSDs replicate
data across the cluster. Consider starting with a 10Gbps network in your racks.
Replicating 1TB of data across a 1Gbps network takes 3 hours, and 3TBs (a
typical drive configuration) takes 9 hours. By contrast, with a 10Gbps network,
the replication times would be 20 minutes and 1 hour respectively. In a
petabyte-scale cluster, failure of an OSD disk should be an expectation, not an
exception. System administrators will appreciate PGs recovering from a
`` degraded `` state to an `` active + clean `` state as rapidly as possible, with
2012-09-27 23:50:16 +00:00
price / performance tradeoffs taken into consideration. Additionally, some
deployment tools (e.g., Dell's Crowbar) deploy with five different networks,
but employ VLANs to make hardware and network cabling more manageable. VLANs
using 802.1q protocol require VLAN-capable NICs and Switches. The added hardware
expense may be offset by the operational cost savings for network setup and
2016-08-05 07:47:36 +00:00
maintenance. When using VLANs to handle VM traffic between the cluster
2012-09-27 23:50:16 +00:00
and compute stacks (e.g., OpenStack, CloudStack, etc.), it is also worth
considering using 10G Ethernet. Top-of-rack routers for each network also need
to be able to communicate with spine routers that have even faster
throughput--e.g., 40Gbps to 100Gbps.
Your server hardware should have a Baseboard Management Controller (BMC).
Administration and deployment tools may also use BMCs extensively, so consider
the cost/benefit tradeoff of an out-of-band network for administration.
Hypervisor SSH access, VM image uploads, OS image installs, management sockets,
etc. can impose significant loads on a network. Running three networks may seem
like overkill, but each traffic path represents a potential capacity, throughput
and/or performance bottleneck that you should carefully consider before
deploying a large scale data cluster.
2012-09-17 23:25:54 +00:00
Failure Domains
===============
A failure domain is any failure that prevents access to one or more OSDs. That
could be a stopped daemon on a host; a hard disk failure, an OS crash, a
malfunctioning NIC, a failed power supply, a network outage, a power outage, and
so forth. When planning out your hardware needs, you must balance the
temptation to reduce costs by placing too many responsibilities into too few
failure domains, and the added costs of isolating every potential failure
domain.
2012-07-10 15:14:42 +00:00
Minimum Hardware Recommendations
================================
Ceph can run on inexpensive commodity hardware. Small production clusters
and development clusters can run successfully with modest hardware.
2013-10-23 00:39:35 +00:00
+--------------+----------------+-----------------------------------------+
| Process | Criteria | Minimum Recommended |
+==============+================+=========================================+
| `` ceph-osd `` | Processor | - 1x 64-bit AMD-64 |
| | | - 1x 32-bit ARM dual-core or better |
| +----------------+-----------------------------------------+
| | RAM | ~1GB for 1TB of storage per daemon |
| +----------------+-----------------------------------------+
| | Volume Storage | 1x storage drive per daemon |
| +----------------+-----------------------------------------+
| | Journal | 1x SSD partition per daemon (optional) |
| +----------------+-----------------------------------------+
| | Network | 2x 1GB Ethernet NICs |
+--------------+----------------+-----------------------------------------+
2016-07-13 06:37:58 +00:00
| `` ceph-mon `` | Processor | - 1x 64-bit AMD-64 |
2013-10-23 00:39:35 +00:00
| | | - 1x 32-bit ARM dual-core or better |
| +----------------+-----------------------------------------+
| | RAM | 1 GB per daemon |
| +----------------+-----------------------------------------+
| | Disk Space | 10 GB per daemon |
| +----------------+-----------------------------------------+
| | Network | 2x 1GB Ethernet NICs |
+--------------+----------------+-----------------------------------------+
| `` ceph-mds `` | Processor | - 1x 64-bit AMD-64 quad-core |
| | | - 1x 32-bit ARM quad-core |
| +----------------+-----------------------------------------+
| | RAM | 1 GB minimum per daemon |
| +----------------+-----------------------------------------+
| | Disk Space | 1 MB per daemon |
| +----------------+-----------------------------------------+
| | Network | 2x 1GB Ethernet NICs |
+--------------+----------------+-----------------------------------------+
2012-07-12 00:35:01 +00:00
2012-11-07 22:09:30 +00:00
.. tip :: If you are running an OSD with a single disk, create a
2012-07-12 00:35:01 +00:00
partition for your volume storage that is separate from the partition
containing the OS. Generally, we recommend separate disks for the
OS and the volume storage.
2012-09-17 23:25:54 +00:00
2013-10-23 00:11:07 +00:00
Production Cluster Examples
===========================
2012-07-12 00:35:01 +00:00
Production clusters for petabyte scale data storage may also use commodity
hardware, but should have considerably more memory, processing power and data
storage to account for heavy traffic loads.
2013-10-23 00:11:07 +00:00
Dell Example
------------
2012-07-12 00:35:01 +00:00
A recent (2012) Ceph cluster project is using two fairly robust hardware
configurations for Ceph OSDs, and a lighter configuration for monitors.
+----------------+----------------+------------------------------------+
| Configuration | Criteria | Minimum Recommended |
+================+================+====================================+
2012-09-07 03:31:46 +00:00
| Dell PE R510 | Processor | 2x 64-bit quad-core Xeon CPUs |
2012-07-12 00:35:01 +00:00
| +----------------+------------------------------------+
| | RAM | 16 GB |
| +----------------+------------------------------------+
2012-09-07 03:31:46 +00:00
| | Volume Storage | 8x 2TB drives. 1 OS, 7 Storage |
2012-07-12 00:35:01 +00:00
| +----------------+------------------------------------+
2012-09-07 03:31:46 +00:00
| | Client Network | 2x 1GB Ethernet NICs |
2012-07-12 00:35:01 +00:00
| +----------------+------------------------------------+
2012-09-07 03:31:46 +00:00
| | OSD Network | 2x 1GB Ethernet NICs |
2012-07-12 00:35:01 +00:00
| +----------------+------------------------------------+
2012-09-27 23:50:16 +00:00
| | Mgmt. Network | 2x 1GB Ethernet NICs |
2012-07-12 00:35:01 +00:00
+----------------+----------------+------------------------------------+
2012-09-07 03:31:46 +00:00
| Dell PE R515 | Processor | 1x hex-core Opteron CPU |
2012-07-12 00:35:01 +00:00
| +----------------+------------------------------------+
| | RAM | 16 GB |
| +----------------+------------------------------------+
2012-09-07 03:31:46 +00:00
| | Volume Storage | 12x 3TB drives. Storage |
2012-07-12 00:35:01 +00:00
| +----------------+------------------------------------+
2012-09-07 03:31:46 +00:00
| | OS Storage | 1x 500GB drive. Operating System. |
2012-07-12 00:35:01 +00:00
| +----------------+------------------------------------+
2012-09-07 03:31:46 +00:00
| | Client Network | 2x 1GB Ethernet NICs |
2012-07-12 00:35:01 +00:00
| +----------------+------------------------------------+
2012-09-07 03:31:46 +00:00
| | OSD Network | 2x 1GB Ethernet NICs |
2012-07-12 00:35:01 +00:00
| +----------------+------------------------------------+
2012-09-27 23:50:16 +00:00
| | Mgmt. Network | 2x 1GB Ethernet NICs |
2012-07-12 00:35:01 +00:00
+----------------+----------------+------------------------------------+
2013-02-01 00:34:02 +00:00
2014-01-17 01:00:38 +00:00
2013-10-23 00:11:07 +00:00
2018-07-17 10:13:58 +00:00
.. _Ceph blog: https://ceph.com/community/blog/
2013-02-01 00:34:02 +00:00
.. _Ceph Write Throughput 1: http://ceph.com/community/ceph-performance-part-1-disk-controller-write-throughput/
.. _Ceph Write Throughput 2: http://ceph.com/community/ceph-performance-part-2-write-throughput-without-ssd-journals/
2017-03-15 08:24:39 +00:00
.. _Mapping Pools to Different Types of OSDs: ../../rados/operations/crush-map#placing-different-pools-on-different-osds
2014-05-02 00:04:43 +00:00
.. _OS Recommendations: ../os-recommendations
