What the sandboxes, volumes and snapshots of the key's project used and kept
The node's GET /v1/usage (sandboxd/openapi/openapi.json) as the control plane answers it. Given none of from, to, group_by, metadata.<key>, kind, volume_id and snapshot_id, it is the node's list of sandboxes (UsageList), asked of every node and put together. Given any of them, it is a report (UsageReport) from the books the control plane keeps by the hour, a --usage-seconds behind the nodes at most and there while a node is away: prices are the fourteen prices, every group and totals have the quantities only its books hold (requests, state_gb_hours, storage_gb_hours, log_bytes, log_gb_hours, egress_eu_bytes, egress_ap_bytes, drive_write_bytes, drive_cold_read_bytes, object_storage_gb_hours, class_a_requests, class_b_requests and object_egress_bytes), and bucket_seconds is 3600. A report by sandbox has one more group, of what no one sandbox used (sandbox_id null): what volumes and snapshots kept and what was written to volumes and read of them cold, the requests apps took and what they printed, what the export of the apps' logs sent (kind log_export), what the backups of volumes kept in the object store (kind backup, in state_gb_hours), and what buckets held, the requests to them and what was read out of them (kind bucket). An account that owes may read it. Takes read.
https://api.glac.ai/v1/usageAuthorization
Query parameters
- group_byoptionalstring
What the groups are of, separated by commas:
sandbox,hourorday,metadata.<key>, and herekind,volume(ordrive: a Drive is a volume, and the groups saydrive_id),snapshotandbackup(each backup of a volume a group of its own,backup_id). Without it a report is by sandbox, by volume when it asks of a volume, by snapshot when it asks of a snapshot, by backup when it asks of a backup (backup_id). Ordatabase(each database of the project a group: what the instances and the Drives of its branches used, its history among them;database_id) orbranch(each branch of a database a group,database_idandbranch_id), withhourordayor alone, and nothing else (slice DB7): what no database's thing used is the group whosedatabase_idis null. - kindoptionalsandboxorvolumeordriveorsnapshotorlog_exportorbackup+1 more
What one kind of thing used or kept:
sandbox,volume(what volumes kept, and what was written to them and read of them cold),drive(the same, said of Drives: a Drive is a volume) orsnapshot;log_export, what the nodes sent the destinations of the logs' export (/v1/destinations), which is booked under each destination (dst-...) and the app; orbackup, what the backups of volumes kept in the object store (/v1/backups), which is booked under each backup (bak-...) on the meter of state (state_gb_hours): itsstored_bytesfrom itsready_atto its delete, or its expiry where that came first, each hour to the hour it was kept in, up to the moment, the hour that has begun among it (booked every minute; one that is deleted, to its delete); orbucket, what the buckets of Object Storage held (object_storage_gb_hours), the requests to them of each class (class_a_requests,class_b_requests) and what was read out of them to the internet (object_egress_bytes, which costs nothing), as their gateway told it (/v1/buckets), booked under each bucket (bkt-...). Another word is 400invalid_field. The control plane's: a node that stands alone refuses it, as it doesvolume_idandsnapshot_id. - backup_idoptionalstring
What one backup of a volume kept in the object store (
/v1/backups): its id (bak-…), whatever became of its volume, on the meter of state (state_gb_hours). One ofsandbox_id,volume_id,drive_id,snapshot_idandbackup_idat a time; grouped by backup unlessgroup_bysays otherwise. The control plane's: a node that stands alone refuses it.
Returns
200
The node's UsageList, or its UsageReport with the control plane's quantities in it (sandboxd/openapi/openapi.json).
One of the following:
- UsageList
- dataarray ofobjectrequired
- sandbox_idstringrequiredpattern: ^sbx-[0-9a-hjkmnp-tv-z]{16}$
- namestring | nullrequired
The sandbox's name; for a deleted one, the name it had last.
- metadatamap ofstringrequired
Free key-value pairs for the caller's own use; lists of sandboxes filter by them. At most 64 keys of 1 to 64 characters, values of up to 1024. Not a place for secrets: metadata are stored in clear, and every set of them a sandbox has had while it used something is kept with its usage for thirteen months, also after the sandbox is deleted (
GET /v1/usagegroups and filters by them). - cpu_secondsnumberrequiredminimum: 0
- memory_gib_secondsnumberrequired
Memory in 128 MiB steps, integrated over time: GiB times seconds.
minimum: 0 - disk_bytesintegerrequired
What the root disk takes now; 0 for a deleted sandbox.
format: int64 · minimum: 0 - egress_bytesintegerrequired
Bytes it has sent to the internet. For a deleted sandbox, less its last second, which is only in the account's month.
format: int64 · minimum: 0 - created_atstringrequiredformat: date-time
- deleted_atstring | nullrequired
Null for a sandbox that exists.
format: date-time - productoptionalbrowsersordatabasesornull
What the sandbox is sold as, by the control plane's own record (never its metadata):
browsersfor a Browsers session (POST /v1/browsers),databasesfor an instance of a database's branch (POST /v1/databases), null for any other sandbox.
- next_cursorstring | nullrequired
Pass it as
cursorto get the next page. Null on the last page. - totalsobjectrequired
Sums over every row the filter selects, not over this page.
- cpu_secondsnumberrequiredminimum: 0
- memory_gib_secondsnumberrequiredminimum: 0
- egress_bytesintegerrequiredformat: int64 · minimum: 0
- sandboxesintegerrequiredminimum: 0
- egressobjectrequired
The account's month: what the allowance is measured against, whichever sandboxes sent it.
- periodstringrequired
The calendar month it is, in UTC.
pattern: ^[0-9]{4}-[0-9]{2}$ - fromstringrequired
The stretch the allowance is for: the period of the account's subscription it is in, or the calendar month.
format: date-time - tostringrequiredformat: date-time
- bytesintegerrequired
Sent in the stretch.
format: int64 · minimum: 0 - allowance_bytesinteger | nullrequired
What the stretch includes; null: whatever is sent is included.
format: int64 · minimum: 0 - overage_bytesinteger | nullrequired
How far
bytesis past the allowance; null without one.format: int64 · minimum: 0 - overage_usdnumberrequired
What the bytes past the allowance come to, at the list price: on a paid plan the period's credit pays it first (
GET /v1/billing,credit), and the account's balance pays what the credit does not.minimum: 0 - regionsobjectrequired
Of the traffic of the stretch, region by region: the region of the node it left from:
naNorth America (and a node with no region),euEurope,apAsia-Pacific. The allowance takes the dearest region's bytes first; each region is priced and rounded on its own, andoverage_usdis theirs added up.- naobjectrequired
- bytesintegerrequired
Sent from the region's nodes in the stretch.
format: int64 · minimum: 0 - overage_bytesinteger | nullrequired
What of them is past the allowance; null without one.
format: int64 · minimum: 0 - unit_price_usdnumberrequired
The region's price of a gigabyte.
- overage_usdnumberrequired
What
overage_bytescomes to at that price, to the nearest millionth.minimum: 0
- euobjectrequired
- bytesintegerrequired
Sent from the region's nodes in the stretch.
format: int64 · minimum: 0 - overage_bytesinteger | nullrequired
What of them is past the allowance; null without one.
format: int64 · minimum: 0 - unit_price_usdnumberrequired
The region's price of a gigabyte.
- overage_usdnumberrequired
What
overage_bytescomes to at that price, to the nearest millionth.minimum: 0
- apobjectrequired
- bytesintegerrequired
Sent from the region's nodes in the stretch.
format: int64 · minimum: 0 - overage_bytesinteger | nullrequired
What of them is past the allowance; null without one.
format: int64 · minimum: 0 - unit_price_usdnumberrequired
The region's price of a gigabyte.
- overage_usdnumberrequired
What
overage_bytescomes to at that price, to the nearest millionth.minimum: 0
- month_bytesintegerrequired
Sent in the calendar month
period: the account's own cap is held against this.format: int64 · minimum: 0 - stoppedbooleanrequired
Whether the account's traffic to the internet is stopped now.
- stopped_reasonstring | nullrequired
allowance_reached: the account has sent its allowance and holds no balance to pay for more;overage_cap_reached: it has sent as much in the calendar month as its own cap. Null when it is not stopped. - pricesobjectrequired
- egress_gb_usdnumberrequired
US dollars for a gigabyte (10^9 bytes) past the allowance.
minimum: 0 - egress_eu_gb_usdnumberrequired
- egress_ap_gb_usdnumberrequired
- provisionalbooleanrequired
- project_bytesintegerrequired
What the project of the calling key sent in the calendar month; the allowance is the account's.
format: int64 · minimum: 0 - countedbooleanrequired
Whether every node asked is counting traffic.
- UsageReport
- dataarray ofobjectrequired
- cpu_secondsnumberrequired
A whole number of microseconds, written in seconds.
minimum: 0 - memory_gib_secondsnumberrequired
Memory in 128 MiB steps, integrated over time: a whole number of steps times milliseconds, written as GiB times seconds (a multiple of 0.000125).
minimum: 0 - egress_bytesintegerrequired
Bytes sent to the internet.
format: int64 · minimum: 0 - requestsintegerrequired
Through a control plane only (a node does not count them): requests into the sandboxes through the platform's ingress that a node took, an HTTP request, a WebSocket and an SSH session each one; what the platform refused itself is none. An app's requests are booked to the app itself, under the labels its instances carry and the region of the edge they came by: grouped by
metadata.appormetadata.regionthey are with what the instances used, and grouped by sandbox they are in the group of no sandbox. control/README.md, 「套餐和账单」.format: int64 · minimum: 0 - state_gb_hoursnumberrequired
Through a control plane only: what the sandboxes kept of their own memory and disk while they slept, sampled once an hour; and what the backups of volumes kept in the object store, the bytes the store holds of each (
stored_bytes) for as long as it is ready and not gone, booked under the backup's own id (kindbackup: 0 in a group of a sandbox). In GB-hours (a whole number of byte-hours, written as 10^-9 of a GB-hour).minimum: 0 - storage_gb_hoursnumberrequired
Through a control plane only: what volumes and snapshots kept, in GB-hours (a whole number of byte-hours, written as 10^-9 of a GB-hour). A volume keeps its
used_bytesand a snapshot itssize_bytes, for as long as each is there, counted by its node from the moment it is made to the moment it is deleted. It is booked under the volume's or the snapshot's own id: 0 in a group of a sandbox.minimum: 0 - log_bytesintegerrequired
Through a control plane only: what apps' instances printed that their nodes kept, in bytes of the lines'
textas they were printed. It is booked to the app itself, in the hour of the lines, under the app's own labels and its id (metadata.app), from what each node counted of the app's log: grouped bymetadata.appit is with what the app's instances used, and grouped by sandbox it is in the group of no sandbox. No meter's: the logs are priced bylog_gb_hours.format: int64 · minimum: 0 - log_gb_hoursnumberrequired
Through a control plane only: what apps' instances printed that their nodes kept, times the hours it was kept (each hour from the one it was printed in to the end of the days the account's plan keeps logs), in GB-hours: the logs (
meters.logsofGET /v1/billing). Booked to the app aslog_bytesis, to the hours it was kept in.minimum: 0 - drive_write_bytesintegerrequired
Through a control plane only: what was written to volumes, as their nodes count it, booked under the volume's own id (0 in a group of a sandbox).
format: int64 · minimum: 0 - drive_cold_read_bytesintegerrequired
Through a control plane only: what was read of volumes from the object store (not from their node's disk), booked under the volume's own id; and what sandboxes read from it through their read-only mounts of Drives' snapshots (slice D7), booked under each sandbox, whose account pays for it.
format: int64 · minimum: 0 - object_storage_gb_hoursnumberrequired
Through a control plane only: what buckets held, in GB-hours, booked under the bucket's own id (
kindbucket).minimum: 0 - class_a_requestsintegerrequired
Through a control plane only: requests to buckets of class A.
format: int64 · minimum: 0 - class_b_requestsintegerrequired
Through a control plane only: requests to buckets of class B.
format: int64 · minimum: 0 - object_egress_bytesintegerrequired
Through a control plane only: what was read out of buckets to the internet, which costs nothing (fair use: control/README.md, 对象存储). It is no part of
egress_bytes.format: int64 · minimum: 0 - browser_awake_secondsnumberrequired
Through a control plane only: how long Browsers sessions were awake, in seconds to the millisecond, as their nodes count it: the meter of the sessions (
browsers), booked of a session and of no other sandbox. A session'scpu_secondsare what of its CPU time was past half of its time awake, hour by hour (counted once each hour is over), and it has nomemory_gib_seconds: asleep, held paused in memory (the pause tier) or not, it has thestate_gb_hoursof what it keeps, as a sleeping sandbox has.minimum: 0 - egress_eu_bytesintegerrequired
Through a control plane only: of
egress_bytes, what left from a node in Europe, which is priced ategress_eu_gb_usd.format: int64 · minimum: 0 - egress_ap_bytesintegerrequired
Through a control plane only: of
egress_bytes, what left from a node in Asia-Pacific, priced ategress_ap_gb_usd; the rest left from one in North America or with no region, priced ategress_gb_usd.format: int64 · minimum: 0 - startoptionalstring
With
hourorday: when the hour or the day (UTC) of this group begins.format: date-time - sandbox_idoptionalstring | null
With
sandbox. Through a control plane, null in the one group of what no one sandbox used: what volumes and snapshots kept (and, where the report is also by volume or by snapshot, in every group of one of those), what the backups of volumes kept, and the requests that apps took, which are booked to the app and to none of its instances.pattern: ^sbx-[0-9a-hjkmnp-tv-z]{16}$ - volume_idoptionalstring | null
Through a control plane only, with
volume: the volume this group is of, or the bucket of an installation of a Managed Application (ibk-, billed as volume data). Null in a group of what no volume kept.pattern: ^(vol|ibk)-[0-9a-hjkmnp-tv-z]{16}$ - drive_idoptionalstring | null
Through a control plane only, with
drive(ordrive_id,kind=drive): the Drive this group is of. Null in a group of what no Drive kept.pattern: ^vol-[0-9a-hjkmnp-tv-z]{16}$ - snapshot_idoptionalstring | null
Through a control plane only, with
snapshot: the snapshot this group is of. Null in a group of what no snapshot kept.pattern: ^snap-[0-9a-hjkmnp-tv-z]{16}$ - backup_idoptionalstring | null
Through a control plane only, with
backup(orbackup_id): the backup of a volume this group is of. Null in a group of what no backup kept. Itsnameis null: a backup has none.pattern: ^bak-[0-9a-z]+$ - kindoptionalstring
Through a control plane only, with
kind(drivein place ofvolumewhere the question says Drives): the kind of thing this group is of.log_export: what the export of the apps' logs sent their destinations, which is traffic to the internet (egress_bytes) and nothing else.backup: what the backups of volumes kept in the object store, on the meter of state (state_gb_hours) and nothing else.bucket: what the buckets of Object Storage held (object_storage_gb_hours), the requests to them (class_a_requests,class_b_requests) and what was read out of them (object_egress_bytes), and nothing else.volume: what volumes kept (storage_gb_hours), and what was written to them and read of them cold (drive_write_bytes,drive_cold_read_bytes). - nameoptionalstring | null
With
sandbox: what the sandbox is called now; for a deleted one, what it was called last. Through a control plane, withvolumeorsnapshot, the same of the volume or the snapshot. - productoptionalbrowsersordatabasesornull
What the sandbox is sold as, by the control plane's own record (never its metadata):
browsersfor a Browsers session (POST /v1/browsers),databasesfor an instance of a database's branch (POST /v1/databases), null for any other sandbox. Through a control plane only, withsandbox. - database_idoptionalstring | null
Through a control plane only, with
sandboxorvolume(drive): the database the thing of this group is of (slice DB7): an instance of one of its branches, or a branch's Drive, by the databases' own records (never metadata). Null for any other thing, and in a group of no thing. Withdatabaseorbranch: the database of this group, null in the group of what no database's thing used.pattern: ^db-[0-9a-hjkmnp-tv-z]{16}$ - branch_idoptionalstring | null
With
database_id: the branch of the database the thing is of (the root's,main, or one of its own). Null wheredatabase_idis. Withbranch: the branch of this group.pattern: ^br-[0-9a-hjkmnp-tv-z]{16}$ - metadataoptionalmap ofstring | null
With
metadata.<key>: each key grouped by, with the value the sandboxes of this group had for it while they used this. Null: they had no such key.
- next_cursorstring | nullrequired
Pass it as
cursorto get the next page. Null on the last page. - totalsobjectrequired
Sums over every group the call selects, not over this page, and how many sandboxes they are of (an app that requests are booked to is not one of them). They are of the moment this page was made. Over a range that includes the present, usage is still being booked: a later page's totals are ahead of an earlier page's, and the groups collected over several pages add up to less than the totals of the last. Over a range that has ended, every page says the same.
- cpu_secondsnumberrequiredminimum: 0
- memory_gib_secondsnumberrequiredminimum: 0
- egress_bytesintegerrequiredformat: int64 · minimum: 0
- requestsintegerrequired
Through a control plane only, as in a group.
format: int64 · minimum: 0 - state_gb_hoursnumberrequired
Through a control plane only, as in a group.
minimum: 0 - storage_gb_hoursnumberrequired
Through a control plane only, as in a group.
minimum: 0 - log_bytesintegerrequired
Through a control plane only, as in a group.
format: int64 · minimum: 0 - log_gb_hoursnumberrequired
Through a control plane only, as in a group.
minimum: 0 - drive_write_bytesintegerrequired
Through a control plane only, as in a group.
format: int64 · minimum: 0 - drive_cold_read_bytesintegerrequired
Through a control plane only, as in a group.
format: int64 · minimum: 0 - object_storage_gb_hoursnumberrequired
Through a control plane only, as in a group.
minimum: 0 - class_a_requestsintegerrequired
Through a control plane only, as in a group.
format: int64 · minimum: 0 - class_b_requestsintegerrequired
Through a control plane only, as in a group.
format: int64 · minimum: 0 - object_egress_bytesintegerrequired
Through a control plane only, as in a group.
format: int64 · minimum: 0 - browser_awake_secondsnumberrequired
Through a control plane only, as in a group.
minimum: 0 - egress_eu_bytesintegerrequired
Through a control plane only, as in a group: of
egress_bytes, what left from a node in Europe.format: int64 · minimum: 0 - egress_ap_bytesintegerrequired
Through a control plane only, as in a group: of
egress_bytes, what left from a node in Asia-Pacific.format: int64 · minimum: 0 - sandboxesintegerrequired
How many sandboxes the groups are of. Volumes and snapshots are not counted here.
minimum: 0
- fromstring | nullrequired
The beginning of the range as it holds: the first hour that begins at or after the
fromasked for. Null: none was asked for.format: date-time - tostring | nullrequired
The end of the range as it holds: the first hour that begins at or after the
toasked for; hours before it are in the range. Null: none was asked for.format: date-time - group_byarray ofstringrequired
What the groups are of, in the order of their keys:
hourorday,sandbox, then eachmetadata.<key>. Through a control planevolume,snapshotandkindcome aftersandbox. - bucket_secondsintegerrequired
The grain of the books: 3600, an hour. (A node started for a test with
--usage-bucket-secondshas a shorter one.)minimum: 1 - metadata_atstringrequired
Which metadata the groups and the filters go by.
time_of_use: the ones a sandbox had while it used what a row holds. - metadata_approximatebooleanrequired
True when something in the range (of the sandbox asked for, whatever the filters by metadata) may be booked under metadata other than the ones it was used under;
metadata_approximate_reasonssays why. The quantities are exact in every case; which group they are in is not, and a filter by metadata may leave some of them out. - metadata_approximate_reasonsarray ofstringrequired
Why, when
metadata_approximateis true: every reason that holds of something in the range, each once; empty when it is false.changed_often: a sandbox's metadata changed more than 60 times in one hour. The first 60 changes each end a row of the books; after that the hour's row goes on under the newest metadata. The others are a control plane's, which keeps the books of many nodes in one store.sets_in_an_hour: one sandbox has rows under at most 61 sets of metadata in an hour, and what it used under a further set is booked under no metadata.values_of_a_key: one metadata key had more distinct values in a calendar month than the books take in for a project (1,000). A value past that takes its key out of the set, and the use is booked under the rest of the set: a sandbox tagged with a team and the id of its job is booked under its team. Asked by that key, such use is under no value; asked by any other key, it is where it belongs.keys_in_a_month: the month's new sets already use 256 distinct key names; a further key is removed from a new set, preserving the other keys and all quantities. Existing sets remain exact.set_too_large: a set of metadata whose JSON text is longer than 4,096 bytes is not kept, and what is used under it is booked under no metadata.sets_in_a_month: the project had more new sets of metadata in a calendar month than the books take in (10,000); what is used under a further new set is booked under no metadata until the month turns. A set the books already have does not count again, and takes no place among a key's values either.sets_in_all: the project keeps as many sets of metadata as the books keep for one in all (130,000: a set is kept for thirteen months after its last use); what is used under a further new set is booked under no metadata until sets the project no longer uses have gone.booked_on_arrival: the sandbox's node gave a reading of the use but no hour and no metadata for it (a node that was cut off before it wrote them down, or one of an earlier version): it is booked in the hour the reading arrived in, under the metadata the sandbox had before or after.unspecified: the hour was marked by a version that did not keep the reason. - pricesobjectrequired
What the quantities are multiplied by: the prices of the account's month, one a meter (and of traffic one a region).
- vcpu_hour_usdnumberrequired
A busy vCPU-hour.
- memory_gib_hour_usdnumberrequired
A GiB-hour of memory held while awake, or paused in memory (the pause tier).
- requests_million_usdnumberrequired
A million requests: into apps, and into sandboxes through their preview URLs.
- state_gb_month_usdnumberrequired
A GB-month of what sleeping sandboxes keep, and of what the backups of volumes keep in the object store.
- egress_gb_usdnumberrequired
A gigabyte (10^9 bytes) sent to the internet from a node in North America, or from one with no region: traffic is priced by the region of the node it left from:
naNorth America (and a node with no region),euEurope,apAsia-Pacific. - egress_eu_gb_usdnumberrequired
A gigabyte sent to the internet from a node in Europe.
- egress_ap_gb_usdnumberrequired
A gigabyte sent to the internet from a node in Asia-Pacific.
- storage_gb_month_usdnumberrequired
A GB-month of what volumes and snapshots keep: the price of what sleeping sandboxes keep, unless the control plane was started with another (
--price-storage-gb-month). - logs_gb_month_usdnumberrequired
A GB-month of what apps print that their nodes keep: the
textof each line of output kept, counted as it was printed, for the hours it is kept (--price-logs-gb-month). - drive_write_gb_usdnumberrequired
A gigabyte (10^9 bytes) written to volumes (Drives): the bytes of the write requests their block devices served (
--price-drive-write-gb). - drive_cold_read_gb_usdnumberrequired
A gigabyte of volumes read cold: fetched from the object store because their node had not got them (a restore from a backup; a block fetched on demand). What is read from the node's own disk costs nothing (
--price-drive-cold-read-gb). - object_storage_gb_month_usdnumberrequired
A GB-month of what buckets hold (
--price-object-storage-gb-month). What is read out of them costs nothing, within fair use. - object_class_a_million_usdnumberrequired
A million requests to buckets of class A: what writes, lists or sets (
--price-object-class-a-million). - object_class_b_million_usdnumberrequired
A million requests to buckets of class B: what reads (
--price-object-class-b-million). What deletes is of neither class and costs nothing. - browser_hour_usdnumberrequired
An hour of a Browsers session awake, by the millisecond, in place of its CPU time and its memory: it includes half a vCPU averaged over each hour it is awake, and its CPU time past that is priced at
vcpu_hour_usd; its memory awake is not priced. Asleep, what it keeps is priced as any sleeping sandbox's (state_gb_month_usd, ormemory_gib_hour_usdwhile it is paused in memory) (--price-browser-hour). - provisionalbooleanrequired
- planstring | nullrequired
The plan the project's account is on now (
Billing.plan). - plan_creditrequired
The credit of the stretch the account is in (
Billing.credit), of all its projects: what is used and left of it so far, beside the quantities that are priced byprices; null on the free plan, which has none. With it a customer checks what a report's quantities come to and what of that the plan's credit pays.}One of the following:
- PlanCredit
- null
400
unknown_parameter; invalid_field: a time that is none, a range that ends before it begins, a group_by or kind that is none, two of sandbox_id, volume_id and snapshot_id.
- moveoptionalobject
With
move_over,move_going_backandmove_taken_over(POST /internal/moves/{id}/cancel): the move, as it stands.- stepstringrequired
Where it stands:
releasing(the old node is asked to release it; the table has it there),taking_over(the new node is asked to take it over; the table has it there),reclaiming(the new node refused it, and the old one is asked to take it back),restoring(the old one kept nothing of it, and takes it over from the object store),done(on the new node),failed(where it was, or back there, or where a failure no time clears or an operator's cancel found it:errorsays why).
- reportoptionalobject
With
compose_refused: the import's report, which says what is refused, why, and what to change.- servicesarray ofobjectrequired
The services the import takes, in the model's order.
- appobjectrequired
The fields of the service's app as the import writes them: what its versions are made of, and what holds at once.
- sizeobjectrequired
What one instance may use at most. It is billed by what it uses.
- memory_bytesintegerrequired
An even number of MiB (a whole number of the 2 MiB pages a node loads a machine's memory in), in bytes. Any other size is 400
invalid_field(paramsize.memory_bytes): no node would make an instance of it, as none makes a sandbox of it (the node'sSandboxLimitsCreate.memory_bytes).minimum: 134217728
- scaleobjectrequired
- min_instancesintegerrequired
How many instances of the app are kept awake in a region, whether a request comes or not: the floor. The nodes of the region that serve the app share it as they share
max_instances(two nodes and a floor of 2: one on each; a floor of 1: one, on the node that is tried first; a floor of 5: three and two), and a node that is away is made up for by the others. A node makes its part when it is told, not at the first request, and neither stops nor deletes those instances for being idle; they are metered as instances that are awake. 0, which is what it is when nothing is said: none. Lowered, the instances beyond it are idle from that moment, and are stopped and deleted as idle instances are. At mostmax_instances, where that is given, and at most what the account's plan lets an app ask for in a region (caps.app_instances.limit: 5, 50 and 500): more is refused, 403plan_cap_reached.minimum: 0 - max_instancesinteger | nullrequired
How many instances the app has at the most in a region: the nodes of the region that serve the app share it. At most what the account's plan lets an app ask for (
caps.app_instances.limit: 5 on the free plan, 50 onp20, 200 onp100, 500 onp300, 1,000 onp1000, 2,000 onp2000): more is refused, 403plan_cap_reached, and what an app asked for on a plan that let it is kept, and held to the plan's number.null: the plan's number for an app that names none (caps.app_instances.default: 2, 10, 20, 50, 100 and 200), ormin_instanceswhere that is more. What the nodes are told is no more than the account has room for awake beside everything else of it that is awake, nor less than the app has awake.minimum: 1 - concurrencyinteger | nullrequired
A cap on what one instance has in hand at a time: the requests the edges sent it, and the connections it has open on its project's network (through its service's address, which the node shares out among the instances connection by connection, or to its own). An instance with that many is full, and another is made for more.
null, which is what it is when nothing is said: no cap. A node then adds an instance when the ones it has are near their size in CPU or memory, and an app does not have to guess a number; a service that takes one request at a time gives one.maximum: 1000 · minimum: 1 - sleep_after_msintegerrequired
After how long without a request an instance stops. One that stays stopped is deleted by its node after
delete_after_ms, and the next request that needs an instance gets a new one; the instancesmin_instanceskeeps awake are neither stopped nor deleted.maximum: 600000 · minimum: 0 - sleep_notice_msinteger | nullrequired
How long before an instance stops for being idle the programs connected to its guest's
/run/sbx/awakeare sent the linesleep N, in milliseconds: they flush what they hold, close their connections, or raise a hold, which puts the stop off. The stop comes that long after, whether they answer or not; where nobody is connected, it is not put off. A keyed instance is told as well. 0 or null: not told (the default).maximum: 60000 · minimum: 0 - sleep_hold_max_msintegerrequired
How long the holds a program in an instance raises keep it up after its time idle has run out, in milliseconds: a program connects to
/run/sbx/awakeand sends+before it answers a request whose work goes on after the answer, and-(or closes the connection, or dies) when that work is done; the instance stopssleep_after_msafter the last hold is lowered. Past this time it stops whatever holds it, the programs connected toldsleep N capfirst wheresleep_notice_msasks for a notice. Held, it runs and is metered as running. 0: holds are not heeded. Null: 30000 (the default).maximum: 900000 · minimum: 0 - delete_after_msinteger | nullrequired
After how long idle an instance is deleted, counted from where its stop is counted from: the next request that needs one gets a new instance, a clone of the version's template, without what the deleted one had gathered in its memory and on its disk (keep what is to last in a volume or a database). A second to an hour, the platform's time, which is the most: more is refused (400
invalid_field), and an instance that is to stay warm is one ofmin_instances.null, which is what it is when nothing is said: the hour of the nodes (--app-delete-after-seconds). A change holds for the instances the app has as for the ones made after: each node judges every instance by the app's time as it is, once a second at the least. Not for the app's keyed instances, whichkeys.delete_after_idle_msis for.maximum: 3600000 · minimum: 1000 - suspend_after_msoptionalinteger | null
How long a stopped instance stays paused before its memory is saved and it sleeps for real, in milliseconds: paused, the next request is answered in a millisecond or two and its memory is metered; asleep, it wakes in a few more and only what it keeps is. Its node holds it to the longest it holds any instance paused (ten minutes); the host's memory running short, a slot another machine needs and a sleep somebody asks for put it to sleep sooner. Null (the default): its node's longest. A database's
suspend_after_secondsis its app's.maximum: 86400000 · minimum: 1000 - wake_daily_atoptionalstring | null
A time of day in UTC,
HH:MM, at which the app's keyed instances and its stateful service's instance that rest are woken every day, for work of their own that waits for a wake (a database's maintenance, say); the instances the nodes make and delete by the app's requests are woken by requests alone. The wake is no activity: what runs then is, and once that is done the instance rests again. A stateful service's instance is woken besides for the next run of its PostgreSQL server'spg_cronjobs, whatever this says. Null (the default): none. A database'swake_daily_atis its app's.pattern: ^([01][0-9]|2[0-3]):[0-5][0-9]$
- egressobject | nullrequired
Egress rules as a sandbox has them (
default,allow,deny,http,inject), null when there are none; kept as the calls gave them (aPATCHmerges them key by key, as a node does). They hold at once and make no version: the nodes that serve the app put them on every instance of its group, the ones there are (one that runs for the connections it has too) and the ones made after; a keyed instance does not have them. The control plane checks only that they are an object of those keys of at most 256 KiB: what the rules say is a node's to check, when it applies them, and an instance whose rules its node cannot read is refused everything (the node's log says why).- deny_portsoptionalarray | null
Ports no connection reaches, TCP or UDP, wherever it goes and whatever else the rules say. A Browsers session's always hold 25, 465, 587 and the ports mining pools listen on (3333, 4444, 5555, 7777, 14444, 45700): the control plane adds them to whatever a change of its rules gives.
- runtimeoptional
Its runtime settings (
AppCreate.runtime).null: none.}One of the following:
- Runtime
- ulimitsmap ofRuntimeLimitGivenrequired
- sysctlsmap ofstringrequired
- tmpfsarray ofobjectrequired
- pathstringrequired
- size_bytesinteger | nullrequired
- modeintegerrequired
- dnsobjectrequired
- serversarray ofstringrequired
- searcharray ofstringrequired
- optionsarray ofstringrequired
- shm_size_bytesinteger | nullrequired
- null
401
missing_api_key, invalid_api_key, expired_api_key, invalid_token.
- moveoptionalobject
With
move_over,move_going_backandmove_taken_over(POST /internal/moves/{id}/cancel): the move, as it stands.- stepstringrequired
Where it stands:
releasing(the old node is asked to release it; the table has it there),taking_over(the new node is asked to take it over; the table has it there),reclaiming(the new node refused it, and the old one is asked to take it back),restoring(the old one kept nothing of it, and takes it over from the object store),done(on the new node),failed(where it was, or back there, or where a failure no time clears or an operator's cancel found it:errorsays why).
- reportoptionalobject
With
compose_refused: the import's report, which says what is refused, why, and what to change.- servicesarray ofobjectrequired
The services the import takes, in the model's order.
- appobjectrequired
The fields of the service's app as the import writes them: what its versions are made of, and what holds at once.
- sizeobjectrequired
What one instance may use at most. It is billed by what it uses.
- memory_bytesintegerrequired
An even number of MiB (a whole number of the 2 MiB pages a node loads a machine's memory in), in bytes. Any other size is 400
invalid_field(paramsize.memory_bytes): no node would make an instance of it, as none makes a sandbox of it (the node'sSandboxLimitsCreate.memory_bytes).minimum: 134217728
- scaleobjectrequired
- min_instancesintegerrequired
How many instances of the app are kept awake in a region, whether a request comes or not: the floor. The nodes of the region that serve the app share it as they share
max_instances(two nodes and a floor of 2: one on each; a floor of 1: one, on the node that is tried first; a floor of 5: three and two), and a node that is away is made up for by the others. A node makes its part when it is told, not at the first request, and neither stops nor deletes those instances for being idle; they are metered as instances that are awake. 0, which is what it is when nothing is said: none. Lowered, the instances beyond it are idle from that moment, and are stopped and deleted as idle instances are. At mostmax_instances, where that is given, and at most what the account's plan lets an app ask for in a region (caps.app_instances.limit: 5, 50 and 500): more is refused, 403plan_cap_reached.minimum: 0 - max_instancesinteger | nullrequired
How many instances the app has at the most in a region: the nodes of the region that serve the app share it. At most what the account's plan lets an app ask for (
caps.app_instances.limit: 5 on the free plan, 50 onp20, 200 onp100, 500 onp300, 1,000 onp1000, 2,000 onp2000): more is refused, 403plan_cap_reached, and what an app asked for on a plan that let it is kept, and held to the plan's number.null: the plan's number for an app that names none (caps.app_instances.default: 2, 10, 20, 50, 100 and 200), ormin_instanceswhere that is more. What the nodes are told is no more than the account has room for awake beside everything else of it that is awake, nor less than the app has awake.minimum: 1 - concurrencyinteger | nullrequired
A cap on what one instance has in hand at a time: the requests the edges sent it, and the connections it has open on its project's network (through its service's address, which the node shares out among the instances connection by connection, or to its own). An instance with that many is full, and another is made for more.
null, which is what it is when nothing is said: no cap. A node then adds an instance when the ones it has are near their size in CPU or memory, and an app does not have to guess a number; a service that takes one request at a time gives one.maximum: 1000 · minimum: 1 - sleep_after_msintegerrequired
After how long without a request an instance stops. One that stays stopped is deleted by its node after
delete_after_ms, and the next request that needs an instance gets a new one; the instancesmin_instanceskeeps awake are neither stopped nor deleted.maximum: 600000 · minimum: 0 - sleep_notice_msinteger | nullrequired
How long before an instance stops for being idle the programs connected to its guest's
/run/sbx/awakeare sent the linesleep N, in milliseconds: they flush what they hold, close their connections, or raise a hold, which puts the stop off. The stop comes that long after, whether they answer or not; where nobody is connected, it is not put off. A keyed instance is told as well. 0 or null: not told (the default).maximum: 60000 · minimum: 0 - sleep_hold_max_msintegerrequired
How long the holds a program in an instance raises keep it up after its time idle has run out, in milliseconds: a program connects to
/run/sbx/awakeand sends+before it answers a request whose work goes on after the answer, and-(or closes the connection, or dies) when that work is done; the instance stopssleep_after_msafter the last hold is lowered. Past this time it stops whatever holds it, the programs connected toldsleep N capfirst wheresleep_notice_msasks for a notice. Held, it runs and is metered as running. 0: holds are not heeded. Null: 30000 (the default).maximum: 900000 · minimum: 0 - delete_after_msinteger | nullrequired
After how long idle an instance is deleted, counted from where its stop is counted from: the next request that needs one gets a new instance, a clone of the version's template, without what the deleted one had gathered in its memory and on its disk (keep what is to last in a volume or a database). A second to an hour, the platform's time, which is the most: more is refused (400
invalid_field), and an instance that is to stay warm is one ofmin_instances.null, which is what it is when nothing is said: the hour of the nodes (--app-delete-after-seconds). A change holds for the instances the app has as for the ones made after: each node judges every instance by the app's time as it is, once a second at the least. Not for the app's keyed instances, whichkeys.delete_after_idle_msis for.maximum: 3600000 · minimum: 1000 - suspend_after_msoptionalinteger | null
How long a stopped instance stays paused before its memory is saved and it sleeps for real, in milliseconds: paused, the next request is answered in a millisecond or two and its memory is metered; asleep, it wakes in a few more and only what it keeps is. Its node holds it to the longest it holds any instance paused (ten minutes); the host's memory running short, a slot another machine needs and a sleep somebody asks for put it to sleep sooner. Null (the default): its node's longest. A database's
suspend_after_secondsis its app's.maximum: 86400000 · minimum: 1000 - wake_daily_atoptionalstring | null
A time of day in UTC,
HH:MM, at which the app's keyed instances and its stateful service's instance that rest are woken every day, for work of their own that waits for a wake (a database's maintenance, say); the instances the nodes make and delete by the app's requests are woken by requests alone. The wake is no activity: what runs then is, and once that is done the instance rests again. A stateful service's instance is woken besides for the next run of its PostgreSQL server'spg_cronjobs, whatever this says. Null (the default): none. A database'swake_daily_atis its app's.pattern: ^([01][0-9]|2[0-3]):[0-5][0-9]$
- egressobject | nullrequired
Egress rules as a sandbox has them (
default,allow,deny,http,inject), null when there are none; kept as the calls gave them (aPATCHmerges them key by key, as a node does). They hold at once and make no version: the nodes that serve the app put them on every instance of its group, the ones there are (one that runs for the connections it has too) and the ones made after; a keyed instance does not have them. The control plane checks only that they are an object of those keys of at most 256 KiB: what the rules say is a node's to check, when it applies them, and an instance whose rules its node cannot read is refused everything (the node's log says why).- deny_portsoptionalarray | null
Ports no connection reaches, TCP or UDP, wherever it goes and whatever else the rules say. A Browsers session's always hold 25, 465, 587 and the ports mining pools listen on (3333, 4444, 5555, 7777, 14444, 45700): the control plane adds them to whatever a change of its rules gives.
- runtimeoptional
Its runtime settings (
AppCreate.runtime).null: none.}One of the following:
- Runtime
- ulimitsmap ofRuntimeLimitGivenrequired
- sysctlsmap ofstringrequired
- tmpfsarray ofobjectrequired
- pathstringrequired
- size_bytesinteger | nullrequired
- modeintegerrequired
- dnsobjectrequired
- serversarray ofstringrequired
- searcharray ofstringrequired
- optionsarray ofstringrequired
- shm_size_bytesinteger | nullrequired
- null
403
key_scope: the key does not allow the scope the call takes (the message names it); needs_api_key: the operator's key, or the SSH gateway's; account_held: the account is held for its payments, and every call of its keys but GET /v1/account is refused so, before the scopes are looked at (x-error-codes); installation_key_scope: a key held to an installation of a Managed Application, on a call that is not about that installation's own things. (A sandbox token is passed to its node, which answers what it has no call for with 404 no_such_route.)
- moveoptionalobject
With
move_over,move_going_backandmove_taken_over(POST /internal/moves/{id}/cancel): the move, as it stands.- stepstringrequired
Where it stands:
releasing(the old node is asked to release it; the table has it there),taking_over(the new node is asked to take it over; the table has it there),reclaiming(the new node refused it, and the old one is asked to take it back),restoring(the old one kept nothing of it, and takes it over from the object store),done(on the new node),failed(where it was, or back there, or where a failure no time clears or an operator's cancel found it:errorsays why).
- reportoptionalobject
With
compose_refused: the import's report, which says what is refused, why, and what to change.- servicesarray ofobjectrequired
The services the import takes, in the model's order.
- appobjectrequired
The fields of the service's app as the import writes them: what its versions are made of, and what holds at once.
- sizeobjectrequired
What one instance may use at most. It is billed by what it uses.
- memory_bytesintegerrequired
An even number of MiB (a whole number of the 2 MiB pages a node loads a machine's memory in), in bytes. Any other size is 400
invalid_field(paramsize.memory_bytes): no node would make an instance of it, as none makes a sandbox of it (the node'sSandboxLimitsCreate.memory_bytes).minimum: 134217728
- scaleobjectrequired
- min_instancesintegerrequired
How many instances of the app are kept awake in a region, whether a request comes or not: the floor. The nodes of the region that serve the app share it as they share
max_instances(two nodes and a floor of 2: one on each; a floor of 1: one, on the node that is tried first; a floor of 5: three and two), and a node that is away is made up for by the others. A node makes its part when it is told, not at the first request, and neither stops nor deletes those instances for being idle; they are metered as instances that are awake. 0, which is what it is when nothing is said: none. Lowered, the instances beyond it are idle from that moment, and are stopped and deleted as idle instances are. At mostmax_instances, where that is given, and at most what the account's plan lets an app ask for in a region (caps.app_instances.limit: 5, 50 and 500): more is refused, 403plan_cap_reached.minimum: 0 - max_instancesinteger | nullrequired
How many instances the app has at the most in a region: the nodes of the region that serve the app share it. At most what the account's plan lets an app ask for (
caps.app_instances.limit: 5 on the free plan, 50 onp20, 200 onp100, 500 onp300, 1,000 onp1000, 2,000 onp2000): more is refused, 403plan_cap_reached, and what an app asked for on a plan that let it is kept, and held to the plan's number.null: the plan's number for an app that names none (caps.app_instances.default: 2, 10, 20, 50, 100 and 200), ormin_instanceswhere that is more. What the nodes are told is no more than the account has room for awake beside everything else of it that is awake, nor less than the app has awake.minimum: 1 - concurrencyinteger | nullrequired
A cap on what one instance has in hand at a time: the requests the edges sent it, and the connections it has open on its project's network (through its service's address, which the node shares out among the instances connection by connection, or to its own). An instance with that many is full, and another is made for more.
null, which is what it is when nothing is said: no cap. A node then adds an instance when the ones it has are near their size in CPU or memory, and an app does not have to guess a number; a service that takes one request at a time gives one.maximum: 1000 · minimum: 1 - sleep_after_msintegerrequired
After how long without a request an instance stops. One that stays stopped is deleted by its node after
delete_after_ms, and the next request that needs an instance gets a new one; the instancesmin_instanceskeeps awake are neither stopped nor deleted.maximum: 600000 · minimum: 0 - sleep_notice_msinteger | nullrequired
How long before an instance stops for being idle the programs connected to its guest's
/run/sbx/awakeare sent the linesleep N, in milliseconds: they flush what they hold, close their connections, or raise a hold, which puts the stop off. The stop comes that long after, whether they answer or not; where nobody is connected, it is not put off. A keyed instance is told as well. 0 or null: not told (the default).maximum: 60000 · minimum: 0 - sleep_hold_max_msintegerrequired
How long the holds a program in an instance raises keep it up after its time idle has run out, in milliseconds: a program connects to
/run/sbx/awakeand sends+before it answers a request whose work goes on after the answer, and-(or closes the connection, or dies) when that work is done; the instance stopssleep_after_msafter the last hold is lowered. Past this time it stops whatever holds it, the programs connected toldsleep N capfirst wheresleep_notice_msasks for a notice. Held, it runs and is metered as running. 0: holds are not heeded. Null: 30000 (the default).maximum: 900000 · minimum: 0 - delete_after_msinteger | nullrequired
After how long idle an instance is deleted, counted from where its stop is counted from: the next request that needs one gets a new instance, a clone of the version's template, without what the deleted one had gathered in its memory and on its disk (keep what is to last in a volume or a database). A second to an hour, the platform's time, which is the most: more is refused (400
invalid_field), and an instance that is to stay warm is one ofmin_instances.null, which is what it is when nothing is said: the hour of the nodes (--app-delete-after-seconds). A change holds for the instances the app has as for the ones made after: each node judges every instance by the app's time as it is, once a second at the least. Not for the app's keyed instances, whichkeys.delete_after_idle_msis for.maximum: 3600000 · minimum: 1000 - suspend_after_msoptionalinteger | null
How long a stopped instance stays paused before its memory is saved and it sleeps for real, in milliseconds: paused, the next request is answered in a millisecond or two and its memory is metered; asleep, it wakes in a few more and only what it keeps is. Its node holds it to the longest it holds any instance paused (ten minutes); the host's memory running short, a slot another machine needs and a sleep somebody asks for put it to sleep sooner. Null (the default): its node's longest. A database's
suspend_after_secondsis its app's.maximum: 86400000 · minimum: 1000 - wake_daily_atoptionalstring | null
A time of day in UTC,
HH:MM, at which the app's keyed instances and its stateful service's instance that rest are woken every day, for work of their own that waits for a wake (a database's maintenance, say); the instances the nodes make and delete by the app's requests are woken by requests alone. The wake is no activity: what runs then is, and once that is done the instance rests again. A stateful service's instance is woken besides for the next run of its PostgreSQL server'spg_cronjobs, whatever this says. Null (the default): none. A database'swake_daily_atis its app's.pattern: ^([01][0-9]|2[0-3]):[0-5][0-9]$
- egressobject | nullrequired
Egress rules as a sandbox has them (
default,allow,deny,http,inject), null when there are none; kept as the calls gave them (aPATCHmerges them key by key, as a node does). They hold at once and make no version: the nodes that serve the app put them on every instance of its group, the ones there are (one that runs for the connections it has too) and the ones made after; a keyed instance does not have them. The control plane checks only that they are an object of those keys of at most 256 KiB: what the rules say is a node's to check, when it applies them, and an instance whose rules its node cannot read is refused everything (the node's log says why).- deny_portsoptionalarray | null
Ports no connection reaches, TCP or UDP, wherever it goes and whatever else the rules say. A Browsers session's always hold 25, 465, 587 and the ports mining pools listen on (3333, 4444, 5555, 7777, 14444, 45700): the control plane adds them to whatever a change of its rules gives.
- runtimeoptional
Its runtime settings (
AppCreate.runtime).null: none.}One of the following:
- Runtime
- ulimitsmap ofRuntimeLimitGivenrequired
- sysctlsmap ofstringrequired
- tmpfsarray ofobjectrequired
- pathstringrequired
- size_bytesinteger | nullrequired
- modeintegerrequired
- dnsobjectrequired
- serversarray ofstringrequired
- searcharray ofstringrequired
- optionsarray ofstringrequired
- shm_size_bytesinteger | nullrequired
- null
429
report_in_progress: another report of the same project was still being made after 2 seconds. retryable.
- moveoptionalobject
With
move_over,move_going_backandmove_taken_over(POST /internal/moves/{id}/cancel): the move, as it stands.- stepstringrequired
Where it stands:
releasing(the old node is asked to release it; the table has it there),taking_over(the new node is asked to take it over; the table has it there),reclaiming(the new node refused it, and the old one is asked to take it back),restoring(the old one kept nothing of it, and takes it over from the object store),done(on the new node),failed(where it was, or back there, or where a failure no time clears or an operator's cancel found it:errorsays why).
- reportoptionalobject
With
compose_refused: the import's report, which says what is refused, why, and what to change.- servicesarray ofobjectrequired
The services the import takes, in the model's order.
- appobjectrequired
The fields of the service's app as the import writes them: what its versions are made of, and what holds at once.
- sizeobjectrequired
What one instance may use at most. It is billed by what it uses.
- memory_bytesintegerrequired
An even number of MiB (a whole number of the 2 MiB pages a node loads a machine's memory in), in bytes. Any other size is 400
invalid_field(paramsize.memory_bytes): no node would make an instance of it, as none makes a sandbox of it (the node'sSandboxLimitsCreate.memory_bytes).minimum: 134217728
- scaleobjectrequired
- min_instancesintegerrequired
How many instances of the app are kept awake in a region, whether a request comes or not: the floor. The nodes of the region that serve the app share it as they share
max_instances(two nodes and a floor of 2: one on each; a floor of 1: one, on the node that is tried first; a floor of 5: three and two), and a node that is away is made up for by the others. A node makes its part when it is told, not at the first request, and neither stops nor deletes those instances for being idle; they are metered as instances that are awake. 0, which is what it is when nothing is said: none. Lowered, the instances beyond it are idle from that moment, and are stopped and deleted as idle instances are. At mostmax_instances, where that is given, and at most what the account's plan lets an app ask for in a region (caps.app_instances.limit: 5, 50 and 500): more is refused, 403plan_cap_reached.minimum: 0 - max_instancesinteger | nullrequired
How many instances the app has at the most in a region: the nodes of the region that serve the app share it. At most what the account's plan lets an app ask for (
caps.app_instances.limit: 5 on the free plan, 50 onp20, 200 onp100, 500 onp300, 1,000 onp1000, 2,000 onp2000): more is refused, 403plan_cap_reached, and what an app asked for on a plan that let it is kept, and held to the plan's number.null: the plan's number for an app that names none (caps.app_instances.default: 2, 10, 20, 50, 100 and 200), ormin_instanceswhere that is more. What the nodes are told is no more than the account has room for awake beside everything else of it that is awake, nor less than the app has awake.minimum: 1 - concurrencyinteger | nullrequired
A cap on what one instance has in hand at a time: the requests the edges sent it, and the connections it has open on its project's network (through its service's address, which the node shares out among the instances connection by connection, or to its own). An instance with that many is full, and another is made for more.
null, which is what it is when nothing is said: no cap. A node then adds an instance when the ones it has are near their size in CPU or memory, and an app does not have to guess a number; a service that takes one request at a time gives one.maximum: 1000 · minimum: 1 - sleep_after_msintegerrequired
After how long without a request an instance stops. One that stays stopped is deleted by its node after
delete_after_ms, and the next request that needs an instance gets a new one; the instancesmin_instanceskeeps awake are neither stopped nor deleted.maximum: 600000 · minimum: 0 - sleep_notice_msinteger | nullrequired
How long before an instance stops for being idle the programs connected to its guest's
/run/sbx/awakeare sent the linesleep N, in milliseconds: they flush what they hold, close their connections, or raise a hold, which puts the stop off. The stop comes that long after, whether they answer or not; where nobody is connected, it is not put off. A keyed instance is told as well. 0 or null: not told (the default).maximum: 60000 · minimum: 0 - sleep_hold_max_msintegerrequired
How long the holds a program in an instance raises keep it up after its time idle has run out, in milliseconds: a program connects to
/run/sbx/awakeand sends+before it answers a request whose work goes on after the answer, and-(or closes the connection, or dies) when that work is done; the instance stopssleep_after_msafter the last hold is lowered. Past this time it stops whatever holds it, the programs connected toldsleep N capfirst wheresleep_notice_msasks for a notice. Held, it runs and is metered as running. 0: holds are not heeded. Null: 30000 (the default).maximum: 900000 · minimum: 0 - delete_after_msinteger | nullrequired
After how long idle an instance is deleted, counted from where its stop is counted from: the next request that needs one gets a new instance, a clone of the version's template, without what the deleted one had gathered in its memory and on its disk (keep what is to last in a volume or a database). A second to an hour, the platform's time, which is the most: more is refused (400
invalid_field), and an instance that is to stay warm is one ofmin_instances.null, which is what it is when nothing is said: the hour of the nodes (--app-delete-after-seconds). A change holds for the instances the app has as for the ones made after: each node judges every instance by the app's time as it is, once a second at the least. Not for the app's keyed instances, whichkeys.delete_after_idle_msis for.maximum: 3600000 · minimum: 1000 - suspend_after_msoptionalinteger | null
How long a stopped instance stays paused before its memory is saved and it sleeps for real, in milliseconds: paused, the next request is answered in a millisecond or two and its memory is metered; asleep, it wakes in a few more and only what it keeps is. Its node holds it to the longest it holds any instance paused (ten minutes); the host's memory running short, a slot another machine needs and a sleep somebody asks for put it to sleep sooner. Null (the default): its node's longest. A database's
suspend_after_secondsis its app's.maximum: 86400000 · minimum: 1000 - wake_daily_atoptionalstring | null
A time of day in UTC,
HH:MM, at which the app's keyed instances and its stateful service's instance that rest are woken every day, for work of their own that waits for a wake (a database's maintenance, say); the instances the nodes make and delete by the app's requests are woken by requests alone. The wake is no activity: what runs then is, and once that is done the instance rests again. A stateful service's instance is woken besides for the next run of its PostgreSQL server'spg_cronjobs, whatever this says. Null (the default): none. A database'swake_daily_atis its app's.pattern: ^([01][0-9]|2[0-3]):[0-5][0-9]$
- egressobject | nullrequired
Egress rules as a sandbox has them (
default,allow,deny,http,inject), null when there are none; kept as the calls gave them (aPATCHmerges them key by key, as a node does). They hold at once and make no version: the nodes that serve the app put them on every instance of its group, the ones there are (one that runs for the connections it has too) and the ones made after; a keyed instance does not have them. The control plane checks only that they are an object of those keys of at most 256 KiB: what the rules say is a node's to check, when it applies them, and an instance whose rules its node cannot read is refused everything (the node's log says why).- deny_portsoptionalarray | null
Ports no connection reaches, TCP or UDP, wherever it goes and whatever else the rules say. A Browsers session's always hold 25, 465, 587 and the ports mining pools listen on (3333, 4444, 5555, 7777, 14444, 45700): the control plane adds them to whatever a change of its rules gives.
- runtimeoptional
Its runtime settings (
AppCreate.runtime).null: none.}One of the following:
- Runtime
- ulimitsmap ofRuntimeLimitGivenrequired
- sysctlsmap ofstringrequired
- tmpfsarray ofobjectrequired
- pathstringrequired
- size_bytesinteger | nullrequired
- modeintegerrequired
- dnsobjectrequired
- serversarray ofstringrequired
- searcharray ofstringrequired
- optionsarray ofstringrequired
- shm_size_bytesinteger | nullrequired
- null
503
report_too_long: the report took more than 5 seconds to make; ask for shorter stretches and add them up, or for whole calendar months. retryable.
- moveoptionalobject
With
move_over,move_going_backandmove_taken_over(POST /internal/moves/{id}/cancel): the move, as it stands.- stepstringrequired
Where it stands:
releasing(the old node is asked to release it; the table has it there),taking_over(the new node is asked to take it over; the table has it there),reclaiming(the new node refused it, and the old one is asked to take it back),restoring(the old one kept nothing of it, and takes it over from the object store),done(on the new node),failed(where it was, or back there, or where a failure no time clears or an operator's cancel found it:errorsays why).
- reportoptionalobject
With
compose_refused: the import's report, which says what is refused, why, and what to change.- servicesarray ofobjectrequired
The services the import takes, in the model's order.
- appobjectrequired
The fields of the service's app as the import writes them: what its versions are made of, and what holds at once.
- sizeobjectrequired
What one instance may use at most. It is billed by what it uses.
- memory_bytesintegerrequired
An even number of MiB (a whole number of the 2 MiB pages a node loads a machine's memory in), in bytes. Any other size is 400
invalid_field(paramsize.memory_bytes): no node would make an instance of it, as none makes a sandbox of it (the node'sSandboxLimitsCreate.memory_bytes).minimum: 134217728
- scaleobjectrequired
- min_instancesintegerrequired
How many instances of the app are kept awake in a region, whether a request comes or not: the floor. The nodes of the region that serve the app share it as they share
max_instances(two nodes and a floor of 2: one on each; a floor of 1: one, on the node that is tried first; a floor of 5: three and two), and a node that is away is made up for by the others. A node makes its part when it is told, not at the first request, and neither stops nor deletes those instances for being idle; they are metered as instances that are awake. 0, which is what it is when nothing is said: none. Lowered, the instances beyond it are idle from that moment, and are stopped and deleted as idle instances are. At mostmax_instances, where that is given, and at most what the account's plan lets an app ask for in a region (caps.app_instances.limit: 5, 50 and 500): more is refused, 403plan_cap_reached.minimum: 0 - max_instancesinteger | nullrequired
How many instances the app has at the most in a region: the nodes of the region that serve the app share it. At most what the account's plan lets an app ask for (
caps.app_instances.limit: 5 on the free plan, 50 onp20, 200 onp100, 500 onp300, 1,000 onp1000, 2,000 onp2000): more is refused, 403plan_cap_reached, and what an app asked for on a plan that let it is kept, and held to the plan's number.null: the plan's number for an app that names none (caps.app_instances.default: 2, 10, 20, 50, 100 and 200), ormin_instanceswhere that is more. What the nodes are told is no more than the account has room for awake beside everything else of it that is awake, nor less than the app has awake.minimum: 1 - concurrencyinteger | nullrequired
A cap on what one instance has in hand at a time: the requests the edges sent it, and the connections it has open on its project's network (through its service's address, which the node shares out among the instances connection by connection, or to its own). An instance with that many is full, and another is made for more.
null, which is what it is when nothing is said: no cap. A node then adds an instance when the ones it has are near their size in CPU or memory, and an app does not have to guess a number; a service that takes one request at a time gives one.maximum: 1000 · minimum: 1 - sleep_after_msintegerrequired
After how long without a request an instance stops. One that stays stopped is deleted by its node after
delete_after_ms, and the next request that needs an instance gets a new one; the instancesmin_instanceskeeps awake are neither stopped nor deleted.maximum: 600000 · minimum: 0 - sleep_notice_msinteger | nullrequired
How long before an instance stops for being idle the programs connected to its guest's
/run/sbx/awakeare sent the linesleep N, in milliseconds: they flush what they hold, close their connections, or raise a hold, which puts the stop off. The stop comes that long after, whether they answer or not; where nobody is connected, it is not put off. A keyed instance is told as well. 0 or null: not told (the default).maximum: 60000 · minimum: 0 - sleep_hold_max_msintegerrequired
How long the holds a program in an instance raises keep it up after its time idle has run out, in milliseconds: a program connects to
/run/sbx/awakeand sends+before it answers a request whose work goes on after the answer, and-(or closes the connection, or dies) when that work is done; the instance stopssleep_after_msafter the last hold is lowered. Past this time it stops whatever holds it, the programs connected toldsleep N capfirst wheresleep_notice_msasks for a notice. Held, it runs and is metered as running. 0: holds are not heeded. Null: 30000 (the default).maximum: 900000 · minimum: 0 - delete_after_msinteger | nullrequired
After how long idle an instance is deleted, counted from where its stop is counted from: the next request that needs one gets a new instance, a clone of the version's template, without what the deleted one had gathered in its memory and on its disk (keep what is to last in a volume or a database). A second to an hour, the platform's time, which is the most: more is refused (400
invalid_field), and an instance that is to stay warm is one ofmin_instances.null, which is what it is when nothing is said: the hour of the nodes (--app-delete-after-seconds). A change holds for the instances the app has as for the ones made after: each node judges every instance by the app's time as it is, once a second at the least. Not for the app's keyed instances, whichkeys.delete_after_idle_msis for.maximum: 3600000 · minimum: 1000 - suspend_after_msoptionalinteger | null
How long a stopped instance stays paused before its memory is saved and it sleeps for real, in milliseconds: paused, the next request is answered in a millisecond or two and its memory is metered; asleep, it wakes in a few more and only what it keeps is. Its node holds it to the longest it holds any instance paused (ten minutes); the host's memory running short, a slot another machine needs and a sleep somebody asks for put it to sleep sooner. Null (the default): its node's longest. A database's
suspend_after_secondsis its app's.maximum: 86400000 · minimum: 1000 - wake_daily_atoptionalstring | null
A time of day in UTC,
HH:MM, at which the app's keyed instances and its stateful service's instance that rest are woken every day, for work of their own that waits for a wake (a database's maintenance, say); the instances the nodes make and delete by the app's requests are woken by requests alone. The wake is no activity: what runs then is, and once that is done the instance rests again. A stateful service's instance is woken besides for the next run of its PostgreSQL server'spg_cronjobs, whatever this says. Null (the default): none. A database'swake_daily_atis its app's.pattern: ^([01][0-9]|2[0-3]):[0-5][0-9]$
- egressobject | nullrequired
Egress rules as a sandbox has them (
default,allow,deny,http,inject), null when there are none; kept as the calls gave them (aPATCHmerges them key by key, as a node does). They hold at once and make no version: the nodes that serve the app put them on every instance of its group, the ones there are (one that runs for the connections it has too) and the ones made after; a keyed instance does not have them. The control plane checks only that they are an object of those keys of at most 256 KiB: what the rules say is a node's to check, when it applies them, and an instance whose rules its node cannot read is refused everything (the node's log says why).- deny_portsoptionalarray | null
Ports no connection reaches, TCP or UDP, wherever it goes and whatever else the rules say. A Browsers session's always hold 25, 465, 587 and the ports mining pools listen on (3333, 4444, 5555, 7777, 14444, 45700): the control plane adds them to whatever a change of its rules gives.
- runtimeoptional
Its runtime settings (
AppCreate.runtime).null: none.}One of the following:
- Runtime
- ulimitsmap ofRuntimeLimitGivenrequired
- sysctlsmap ofstringrequired
- tmpfsarray ofobjectrequired
- pathstringrequired
- size_bytesinteger | nullrequired
- modeintegerrequired
- dnsobjectrequired
- serversarray ofstringrequired
- searcharray ofstringrequired
- optionsarray ofstringrequired
- shm_size_bytesinteger | nullrequired
- null