mirror of https://gitee.com/bigwinds/arangodb
improve documentation a bit (#8722)
This commit is contained in:
Jan
GitHub
parent
9c4af6cb7d
commit
44bc625317
|
|
@ -7,39 +7,42 @@ Introduction to TTL (time-to-live) Indexes
|
|||
The TTL index provided by ArangoDB is used for removing expired documents
|
||||
from a collection.
|
||||
|
||||
The TTL index is set up by setting an `expireAfter` value and by picking a single
|
||||
document attribute which contains the documents' creation date and time. Documents
|
||||
are expired after `expireAfter` seconds after their creation time. The creation time
|
||||
is specified as a numeric timestamp (Unix timestamp) or a date string in format
|
||||
`YYYY-MM-DDTHH:MM:SS` with optional milliseconds. All date strings will be interpreted
|
||||
as UTC dates.
|
||||
The TTL index is set up by setting an `expireAfter` value and by selecting a single
|
||||
document attribute which contains a reference timepoint. For each document, that
|
||||
reference timepoint can then be specified as a numeric timestamp (Unix timestamp) or
|
||||
a date string in format `YYYY-MM-DDTHH:MM:SS` with optional milliseconds.
|
||||
All date strings will be interpreted as UTC dates.
|
||||
|
||||
For example, if `expireAfter` is set to 600 seconds (10 minutes) and the index
|
||||
attribute is "creationDate" and there is the following document:
|
||||
Documents will count as expired when wall clock time is beyond the per-document
|
||||
reference timepoint value plus the index' `expireAfter` value added to it.
|
||||
|
||||
### Removing documents at a fixed period after creation / update
|
||||
|
||||
One use case supported by TTL indexes is to remove documents at a fixed duration
|
||||
after they have been created or last updated. This requires setting up the index
|
||||
with an attribute that contains the documents' creation or last-updated time.
|
||||
|
||||
Let's assume the index attribute is set to "creationDate", and the `expireAfter`
|
||||
attribute of the index was set to 600 seconds (10 minutes).
|
||||
|
||||
db.collection.ensureIndex({ type: "ttl", fields: ["creationDate"], expireAfter: 600 });
|
||||
|
||||
Let's further assume the following document now gets inserted into the collection:
|
||||
|
||||
{ "creationDate" : 1550165973 }
|
||||
|
||||
This document will be indexed with a creation date time value of `1550165973`,
|
||||
which translates to the human-readable date `2019-02-14T17:39:33.000Z`. The document
|
||||
This document will be indexed with a reference timepoint value of `1550165973`,
|
||||
which translates to the human-readable date/time `2019-02-14T17:39:33.000Z`. The document
|
||||
will expire 600 seconds afterwards, which is at timestamp `1550166573` (or
|
||||
`2019-02-14T17:49:33.000Z` in the human-readable version).
|
||||
|
||||
The actual removal of expired documents will not necessarily happen immediately.
|
||||
Expired documents will eventually removed by a background thread that is periodically
|
||||
going through all TTL indexes and removing the expired documents.
|
||||
|
||||
There is no guarantee when exactly the removal of expired documents will be carried
|
||||
out, so queries may still find and return documents that have already expired. These
|
||||
will eventually be removed when the background thread kicks in and has capacity to
|
||||
remove the expired documents. It is guaranteed however that only documents which are
|
||||
past their expiration time will actually be removed.
|
||||
`2019-02-14T17:49:33.000Z` in the human-readable version). From that point on, the
|
||||
document is a candidate for being removed.
|
||||
|
||||
Please note that the numeric date time values for the index attribute should be
|
||||
specified in milliseconds since January 1st 1970 (Unix timestamp). To calculate the current
|
||||
specified in seconds since January 1st 1970 (Unix timestamp). To calculate the current
|
||||
timestamp from JavaScript in this format, there is `Date.now() / 1000`, to calculate it
|
||||
from an arbitrary Date instance, there is `Date.getTime() / 1000`.
|
||||
from an arbitrary `Date` instance, there is `Date.getTime() / 1000`.
|
||||
|
||||
Alternatively, the index attribute values can be specified as a date string in format
|
||||
Alternatively, the reference timepoints can be specified as a date string in format
|
||||
`YYYY-MM-DDTHH:MM:SS` with optional milliseconds. All date strings will be interpreted
|
||||
as UTC dates.
|
||||
|
||||
|
|
@ -47,17 +50,61 @@ The above example document using a datestring attribute value would be
|
|||
|
||||
{ "creationDate" : "2019-02-14T17:39:33.000Z" }
|
||||
|
||||
Now any data-modification access to the document could update the value in the document's
|
||||
`creationDate` attribute to the current date/time, which would prolong the existence
|
||||
of the document and keep it from being expired and removed.
|
||||
|
||||
Setting a document's reference timepoint on initial insertion or updating it on every
|
||||
subsequent modification of the document will not be performed by ArangoDB. Instead, it
|
||||
is the tasks of client applications to set and update the reference timepoints whenever
|
||||
the use case requires it.
|
||||
|
||||
### Removing documents at certain points in time
|
||||
|
||||
Another use case is to specify a per-document expiration/removal timepoint, and setting
|
||||
the `expireAfter` attribute to a low value (e.g. 0 seconds).
|
||||
|
||||
Let's assume the index attribute is set to "expireDate", and the `expireAfter`
|
||||
attribute of the index was set to 0 seconds (immediately when wall clock time reaches
|
||||
the value specified in `expireDate`).
|
||||
|
||||
db.collection.ensureIndex({ type: "ttl", fields: ["expireDate"], expireAfter: 0 });
|
||||
|
||||
When storing the following document in the collection, it will expire at the timepoint
|
||||
specified in the document itself:
|
||||
|
||||
{ "expireDate" : "2019-03-28T01:06:00Z" }
|
||||
|
||||
As `expireAfter` was set to 0, the document will count as expired when wall clock time
|
||||
has reached the timeout.
|
||||
|
||||
It should be intuitive to see that the `expireDate` can be differently per document.
|
||||
This allows mixing of documents with different expiration periods by calculating their
|
||||
expiration dates differently in the client application.
|
||||
|
||||
### Preventing documents from being removed
|
||||
|
||||
In case the index attribute does not contain a numeric value nor a proper date string,
|
||||
the document will not be stored in the TTL index and thus will not become a candidate
|
||||
for expiration and removal. Providing either a non-numeric value or even no value for
|
||||
the index attribute is a supported way of keeping documents from being expired and removed.
|
||||
the index attribute is a supported way to keep documents from being expired and removed.
|
||||
|
||||
There can at most be one TTL index per collection. It is not recommended to use
|
||||
TTL indexes for user-land AQL queries, as TTL indexes may store a transformed,
|
||||
always numerical version of the index attribute value.
|
||||
### Limitations
|
||||
|
||||
The actual removal of expired documents will not necessarily happen immediately when
|
||||
they have reached their expiration time.
|
||||
Expired documents will eventually be removed by a background thread that is periodically
|
||||
going through all TTL indexes and removing the expired documents.
|
||||
|
||||
There is no guarantee when exactly the removal of expired documents will be carried
|
||||
out, so queries may still find and return documents that have already expired. These
|
||||
will eventually be removed when the background thread kicks in and has spare capacity to
|
||||
remove the expired documents. It is guaranteed however that only documents which are
|
||||
past their expiration time will actually be removed.
|
||||
|
||||
The frequency for invoking the background removal thread can be configured using
|
||||
the `--ttl.frequency` startup option. The frequency is specified in milliseconds.
|
||||
|
||||
The frequency for invoking the background removal thread can be configured
|
||||
using the `--ttl.frequency` startup option.
|
||||
In order to avoid "random" load spikes by the background thread suddenly kicking
|
||||
in and removing a lot of documents at once, the number of to-be-removed documents
|
||||
per thread invocation can be capped.
|
||||
|
|
@ -66,6 +113,19 @@ controlled by the startup option `--ttl.max-total-removes`. The maximum number o
|
|||
documents in a single collection at once can be controlled by the startup option
|
||||
`--ttl.max-collection-removes`.
|
||||
|
||||
There can at most be one TTL index per collection. It is not recommended to rely on
|
||||
TTL indexes for user-land AQL queries. This is because TTL indexes may store a transformed,
|
||||
always numerical version of the index attribute value even if it was originally passed
|
||||
in as a datestring.
|
||||
|
||||
Please note that there is one background thread per ArangoDB database server instance
|
||||
for performing the removal of expired documents of all collections in all databases.
|
||||
If the number of databases and collections with TTL indexes is high and there are many
|
||||
documents to remove from these, the background thread may at least temporarily lag
|
||||
behind with its removal operations. It should eventually catch up in case the number
|
||||
of to-be-removed documents per invocation is not higher than the background thread's
|
||||
configured threshold values.
|
||||
|
||||
|
||||
Accessing TTL Indexes from the Shell
|
||||
-------------------------------------
|
||||
|
|
|
|||
|
|
@ -65,11 +65,11 @@ different usage scenarios:
|
|||
expired documents from a collection.
|
||||
|
||||
The TTL index is set up by setting an `expireAfter` value and by picking a single
|
||||
document attribute which contains the documents' creation date and time. Documents
|
||||
are expired after `expireAfter` seconds after their creation time. The creation time
|
||||
is specified as either a numeric timestamp (Unix timestamp) or a date string in format
|
||||
`YYYY-MM-DDTHH:MM:SS` with optional milliseconds. All date strings will be interpreted
|
||||
as UTC dates.
|
||||
document attribute which contains the documents' reference timepoint. Documents
|
||||
are expired `expireAfter` seconds after their reference timepoint has been reached.
|
||||
The documents' reference timepoint is specified as either a numeric timestamp
|
||||
(Unix timestamp) or a date string in format `YYYY-MM-DDTHH:MM:SS` with optional
|
||||
milliseconds. All date strings will be interpreted as UTC dates.
|
||||
|
||||
For example, if `expireAfter` is set to 600 seconds (10 minutes) and the index
|
||||
attribute is "creationDate" and there is the following document:
|
||||
|
|
@ -94,7 +94,7 @@ different usage scenarios:
|
|||
past their expiration time will actually be removed.
|
||||
|
||||
Please note that the numeric date time values for the index attribute should be
|
||||
specified in milliseconds since January 1st 1970 (Unix timestamp). To calculate the current
|
||||
specified in seconds since January 1st 1970 (Unix timestamp). To calculate the current
|
||||
timestamp from JavaScript in this format, there is `Date.now() / 1000`, to calculate it
|
||||
from an arbitrary Date instance, there is `Date.getTime() / 1000`.
|
||||
|
||||
|
|
@ -111,8 +111,9 @@ different usage scenarios:
|
|||
for expiration and removal. Providing either a non-numeric value or even no value for
|
||||
the index attribute is a supported way of keeping documents from being expired and removed.
|
||||
|
||||
It is not recommended to use TTL indexes for user-land AQL queries, as TTL indexes may
|
||||
store a transformed, always numerical version of the index attribute value.
|
||||
It is not recommended to rely on TTL indexes for user-land AQL queries. This is because
|
||||
TTL indexe may store a transformed, always numerical version of the index attribute value
|
||||
even if it was originally passed in as a datestring.
|
||||
|
||||
- geo index: the geo index provided by ArangoDB allows searching for documents
|
||||
within a radius around a two-dimensional earth coordinate (point), or to
|
||||
|
|
|
|||
|
|
@ -190,65 +190,18 @@ other operations on the collection.
|
|||
TTL (time-to-live) Indexes
|
||||
--------------------------
|
||||
|
||||
The new TTL indexes provided by ArangoDB can be used for removing expired documents
|
||||
from a collection.
|
||||
The new TTL indexes feature provided by ArangoDB can be used for automatically
|
||||
removing expired documents from a collection.
|
||||
|
||||
A TTL index can be set up by setting an `expireAfter` value and by picking a single
|
||||
document attribute which contains the documents' creation date and time. Documents
|
||||
expire `expireAfter` seconds after their creation time. The creation time
|
||||
is specified as either a numeric timestamp or a UTC datestring.
|
||||
TTL indexes support eventual removal of documents which are past a configured
|
||||
expiration timepoint. The expiration timepoints can be based upon the documents'
|
||||
original insertion or last-updated timepoints, with adding a period during
|
||||
which to retain the documents.
|
||||
Alternatively, expiration timepoints can be specified as absolute values per
|
||||
document.
|
||||
It is also possible to exclude documents from automatic expiration and removal.
|
||||
|
||||
For example, if `expireAfter` is set to 600 seconds (10 minutes) and the index
|
||||
attribute is "creationDate" and there is the following document:
|
||||
|
||||
{ "creationDate" : 1550165973 }
|
||||
|
||||
This document will be indexed with a creation timestamp value of `1550165973`,
|
||||
which translates to the human-readable date string `2019-02-14T17:39:33.000Z`. The
|
||||
document will expire 600 seconds afterwards, which is at timestamp `1550166573` (or
|
||||
`2019-02-14T17:49:33.000Z` in the human-readable version).
|
||||
|
||||
The actual removal of expired documents will not necessarily happen immediately.
|
||||
Expired documents will eventually removed by a background thread that is periodically
|
||||
going through all TTL indexes and removing the expired documents.
|
||||
|
||||
There is no guarantee when exactly the removal of expired documents will be carried
|
||||
out, so queries may still find and return documents that have already expired. These
|
||||
will eventually be removed when the background thread kicks in and has capacity to
|
||||
remove the expired documents. It is guaranteed however that only documents which are
|
||||
past their expiration time will actually be removed.
|
||||
|
||||
Please note that the numeric timestamp values for the index attribute should be
|
||||
specified in seconds since January 1st 1970 (Unix timestamp). To calculate the current
|
||||
timestamp from JavaScript in this format, there is `Date.now() / 1000`, to calculate it
|
||||
from an arbitrary Date instance, there is `Date.getTime() / 1000`.
|
||||
|
||||
Alternatively, the index attribute values can be specified as a date string in format
|
||||
`YYYY-MM-DDTHH:MM:SS` with optional milliseconds. All date strings will be interpreted
|
||||
as UTC dates.
|
||||
|
||||
The above example document using a datestring attribute value would be
|
||||
|
||||
{ "creationDate" : "2019-02-14T17:39:33.000Z" }
|
||||
|
||||
In case the index attribute does not contain a numeric value nor a proper date string,
|
||||
the document will not be stored in the TTL index and thus will not become a candidate
|
||||
for expiration and removal. Providing either a non-numeric value or even no value for
|
||||
the index attribute is a supported way of keeping documents from being expired and removed.
|
||||
|
||||
There can at most be one TTL index per collection. It is not recommended to use
|
||||
TTL indexes for user-land AQL queries, as TTL indexes may store a transformed,
|
||||
always numerical version of the index attribute value.
|
||||
|
||||
The frequency for invoking the background removal thread can be configured
|
||||
using the `--ttl.frequency` startup option.
|
||||
In order to avoid "random" load spikes by the background thread suddenly kicking
|
||||
in and removing a lot of documents at once, the number of to-be-removed documents
|
||||
per thread invocation can be capped.
|
||||
The total maximum number of documents to be removed per thread invocation is
|
||||
controlled by the startup option `--ttl.max-total-removes`. The maximum number of
|
||||
documents in a single collection at once can be controlled by the startup option
|
||||
`--ttl.max-collection-removes`.
|
||||
Also see the [TTL Indexes](../Indexing/Ttl.md) page.
|
||||
|
||||
|
||||
HTTP API extensions
|
||||
|
|
|
|||
|
|
@ -108,6 +108,9 @@ Result TtlProperties::fromVelocyPack(VPackSlice const& slice) {
|
|||
return Result(TRI_ERROR_BAD_PARAMETER, "expecting numeric value for frequency");
|
||||
}
|
||||
frequency = slice.get("frequency").getNumericValue<uint64_t>();
|
||||
if (frequency < TtlProperties::minFrequency) {
|
||||
return Result(TRI_ERROR_BAD_PARAMETER, "too low value for frequency");
|
||||
}
|
||||
}
|
||||
if (slice.hasKey("maxTotalRemoves")) {
|
||||
if (!slice.get("maxTotalRemoves").isNumber()) {
|
||||
|
|
@ -439,6 +442,12 @@ void TtlFeature::validateOptions(std::shared_ptr<ProgramOptions> options) {
|
|||
<< "invalid value for '--ttl.max-collection-removes'.";
|
||||
FATAL_ERROR_EXIT();
|
||||
}
|
||||
|
||||
if (_properties.frequency < TtlProperties::minFrequency) {
|
||||
LOG_TOPIC("ea696", FATAL, arangodb::Logger::STARTUP)
|
||||
<< "too low value for '--ttl.frequency'.";
|
||||
FATAL_ERROR_EXIT();
|
||||
}
|
||||
}
|
||||
|
||||
void TtlFeature::start() {
|
||||
|
|
|
|||
|
|
@ -55,6 +55,7 @@ struct TtlStatistics {
|
|||
};
|
||||
|
||||
struct TtlProperties {
|
||||
static constexpr uint64_t minFrequency = 1 * 1000; // milliseconds
|
||||
uint64_t frequency = 30 * 1000; // milliseconds
|
||||
uint64_t maxTotalRemoves = 1000000;
|
||||
uint64_t maxCollectionRemoves = 1000000;
|
||||
|
|
|
|||
Loading…
Reference in New Issue