Capacity planning
Capacity planning should be considered as part of the requirements of deploying QuestDB to forecast CPU, memory, network capacity, and a combination of these elements, depending on the expected demands of the system. This page describes configuring these system resources with example scenarios that align with both edge cases and common setup configurations.
Most of the configuration settings referred to below except for OS settings are
configured in QuestDB by either a server.conf configuration file or as
environment variables. For more details on applying configuration settings in
QuestDB, refer to the configuration page.
To monitor various metrics of the QuestDB instances, refer to the Prometheus monitoring page or the Health monitoring page.
Storage and filesystem#
The following sections describe aspects to consider regarding the storage of data and file systems.
Supported filesystem#
QuestDB officially supports the following filesystems:
- APFS
- EXT4
- NTFS
- OVERLAYFS (used by Docker)
- XFS
Other file systems supporting mmap feature may work with QuestDB but they should not be used in production, as QuestDB does not run tests on them.
When an unsupported file system is used, QuestDB logs show the following warning:
caution
Users can't use NFS or similar distributed filesystems directly with a QuestDB database.
Write amplification#
In QuestDB the write amplification is calculated by the
metrics:
questdb_physically_written_rows_total / questdb_committed_rows_total.
When ingesting out-of-order data, a high disk write rate combined with high write amplification may slow down the performance.
For data ingestion over PGWire, or as a further step for ILP ingestion, smaller table partitions maybe reduce the write amplification. This applies to tables with partition directories exceeding a few hundred MBs on disk. For example, partition by day can be reduced to by hour, partition by month to by day, and so on.
Partition split#
From QuestDB 7.2, heavily out-of-order commits can split the partitions into parts to reduce write amplification. When data is merged into an existing partition as a result of an out-of-order insert, the partition will be split into two parts: the prefix sub-partition and the suffix sub-partition.
Considering an example of the following partition details:
- A partition
2023-01-01.1with 1,000 rows every hour, and therefore 24,000 rows in total. - Inserting one row with the timestamp
2023-01-01T23:00
When the out-of-order row 2023-01-01T23:00 is inserted, the partition is split
into 2 parts:
- Prefix:
2023-01-01.1with 23,000 rows - Suffix (including the merged row):
2023-01-01T75959-999999.2with 1,001 rows
See Splitting and squashing time partitions for more information.
CPU and RAM configuration#
This section describes configuration strategies based on the forecast behavior of the database.
RAM size#
We recommend having at least 8GB of RAM for basic workloads and 32GB for more advanced ones.
For relatively small datasets, typically a few to a few dozen GB, if the need for reads is high, performance can benefit from maximizing the use of the OS page cache. Users may consider increasing available RAM to improve the speed of read operations.
Memory page size configuration#
For frequent out-of-order (O3) writes over a high number of columns/tables, the
performance may be impacted by the size of the memory page being too big as this
increases the demand for RAM. The memory page, cairo.o3.column.memory.size, is
set to 8M by default. This means that the table writer uses 16MB (2x8MB) RAM per
column when it receives O3 writes. Decreasing the value in the interval of
[128K, 8M] based on the number of columns used may improve O3 write performance.
CPU cores#
By default, QuestDB attempts to use all available CPU cores. The guide on shared worker configuration details how to change the default setting. Assuming that the disk does not have bottleneck for operations, the throughput of read-only queries scales proportionally with the number of available cores. As a result, a machine with more cores will provide better query performance.
Shared workers#
In QuestDB, there are worker pools which can help separate CPU-load between sub-systems.
caution
In case if you are configuring thread pool sizes manually, the total number of threads to be used by QuestDB should not exceed the number of available CPU cores.
The number of worker threads shared across the application can be configured as well as affinity to pin processes to specific CPUs by ID. Shared worker threads service SQL execution subsystems and, in the default configuration, every other subsystem. More information on these settings can be found on the shared worker configuration page.
QuestDB will allocate CPU resources differently depending on how many CPU cores are available. This default can be overridden via configuration. We recommend at least 4 cores for basic workloads and 16 for advanced ones.
8 CPU cores or less#
QuestDB will configure a shared worker pool to handle everything except the InfluxDB line protocol (ILP) writer which gets a dedicated CPU core. The worker count is calculated as follows:
$(cpuAvailable) - (line.tcp.writer.worker.count)$ The minimal size of the shared worker pool is 2, even on a single-core machine.
16 CPU cores or less#
ILP I/O Worker pool is configured to use 2 CPU cores to speed up ingestion and the ILP Writer is using 1 core. The shared worker pool is handling everything else and is configured using this formula:
$(cpuAvailable) - 1 - (line.tcp.writer.worker.count) - (line.tcp.io.worker.count)$
For example, with 16 cores, the shared pool will have 12 threads:
$16-1-2-1$
17 CPU cores and more#
The ILP I/O Worker pool is configured to use 6 CPU cores to speed up ingestion and the ILP Writer is using 1 core. The shared worker pool is handling everything else and is configured using this formula:
$(cpuAvailable) - 2 - (line.tcp.writer.worker.count) - (line.tcp.io.worker.count)$
For example, with 32 cores, the shared pool will have 23 threads:
$32-2-6-1$
Writer page size#
The default page size for writers is 16MB. In cases where there are a large
number of small tables, using 16MB to write a maximum of 1MB of data, for
example, is a waste of OS resources. To change the default value, set the
cairo.writer.data.append.page.size value in server.conf:
InfluxDB over TCP#
We have a documentation page dedicated to capacity planning for ILP ingestion.
InfluxDB over UDP#
note
The UDP receiver is deprecated since QuestDB version 6.5.2. We recommend the TCP receiver instead.
Given a single client sending data to QuestDB via InfluxDB line protocol over UDP, the following configuration can be applied which dedicates a thread for a UDP writer and specifies a CPU core by ID:
Postgres#
Given clients sending data to QuestDB via Postgres interface, the following
configuration can be applied which sets a dedicated worker and pins it with
affinity to a CPU by core ID:
Network Configuration#
For InfluxDB line, PGWire and HTTP protocols, there are a set of configuration
settings relating to the number of clients that may connect, the internal I/O
capacity and connection timeout settings. These settings are configured in the
server.conf file in the format:
Where <protocol> is one of:
http- HTTP connectionspg- PGWire protocolline.tcp- InfluxDB line protocol over TCP
And <config> is one of the following settings:
| key | description |
|---|---|
limit | The number of simultaneous connections to the server. This value is intended to control server memory consumption. |
timeout | Connection idle timeout in milliseconds. Connections are closed by the server when this timeout lapses. |
hint | Applicable only for Windows, where TCP backlog limit is hit. For example Windows 10 allows max of 200 connection. Even if limit is set higher, without hint=true it won't be possible to connect more than 200 connection. |
sndbuf | Maximum send buffer size on each TCP socket. If value is -1 socket send buffer remains unchanged from OS default. |
rcvbuf | Maximum receive buffer size on each TCP socket. If value is -1, the socket receive buffer remains unchanged from OS default. |
For example, this is the configuration for Linux with a relatively low number of concurrent connections:
Let's assume you would like to configure InfluxDB line protocol for a large number of concurrent connections on Windows:
For reference on the defaults of the http and pg protocols, refer to the
server configuration page.
Pooled connection#
Connection pooling should be used for any production-ready use of PGWire or ILP over TCP.
The maximum number of pooled connections is configurable,
(pg.connection.pool.capacity for PGWire and
(line.tcp.connection.pool.capacity for ILP over TCP. The default number of
connections for both interfaces is 64. Users should avoid using too many
connections.
OS configuration#
This section describes approaches for changing system settings on the host QuestDB is running on when system limits are reached due to maximum open files or virtual memory areas. QuestDB passes operating system errors to its logs unchanged and as such, changing the following system settings should only be done in response to such OS errors.
Maximum open files#
QuestDB uses a columnar storage model and
therefore most data structures relate closely to the file system, with columnar
data being stored in its own .d file per partition. In edge cases with
extremely large tables, frequent out-of-order ingestion, or a high number of
table partitions, the number of open files may hit a user or system-wide maximum
limit and can cause unpredictable behavior.
In a Linux/macOS environment, the following commands allow for checking the current user limits for the maximum number of open files:
Setting the open file limit for the current user:#
On a Linux environment, it is enough to increase the hard limit, while on macOS, both hard and soft limits should be set. See Max Open Files Limit on MacOS for the JVM for more details.
Modify user limits using ulimit:
If the user limit is set above the system-wide limit, the system-wide limit should be increased to the same value, too.
Setting the system-wide open file limit on Linux:#
To increase this setting and have the configuration persistent, the limit on the
number of concurrently open files can be changed in /etc/sysctl.conf:
To confirm that this value has been correctly configured, reload sysctl and
check the current value:
Setting system-wide open file limit on macOS:#
On macOS, the system-wide limit can be modified by using launchctl:
To confirm the change, view the current settings using sysctl:
Max virtual memory areas limit#
If the host machine has insufficient limits of map areas, this may result in
out- of-memory exceptions. To increase this value and have the configuration
persistent, mapped memory area limits can be changed in /etc/sysctl.conf:
Each mapped area needs kernel memory, and it's recommended to have around 128 bytes available per 1 map count.