ezgh warehouse warehouses create <name> --engine-version v --instance-type t --storage-size-gib n --subnet s [flags] [--wait]Create a warehouse in a subnet of your VPC. ezgh warehouse warehouses engine-versions lists the versions. --shards (1 to 8) and --replicas (1 to 3) set its shape, 1 × 1 by default; each shard keeps serving through the loss of one node with 2 or 3 replicas. --keeper is dedicated (the default for more than one node) or colocated, for 3 or more nodes of at least 8 vCPUs and 32 GiB. Every node’s instance type needs at least 8 GiB of memory. --from-backup <whs_ ID>/<backup ID> restores a backup into the new warehouse, with the backup’s number of shards as --shards (ezgh warehouse backups get shows it); it needs warehouse.warehouses.query on the backup’s warehouse. Its users come across with new passwords in new Secrets; connections don’t. The app user’s password is written to a new Secret in the project, which ezgh warehouse users get <warehouse> app names. It takes minutes: --wait waits until it’s available and prints it.
Part of ezgh warehouse warehouses.
Examples
ezgh warehouse warehouses create analytics --engine-version 26.8 --shards 2 --replicas 2 \
--instance-type gp-1.large --storage-size-gib 500 --subnet sbn_2m9x0a7qpe4k --waitFlags
| Flag | Type | Default | Description |
|---|---|---|---|
--allowed-source-cidr |
string, repeatable | an IPv4 CIDR that may connect; repeat for more, at most 20 (default the VPC’s ranges) | |
--backup-retention-days |
int | how long backups are kept, 1 to 35 days (default 7) | |
--backup-window |
string | when the daily backup is taken, HH:MM-HH:MM UTC | |
--backups |
boolean | true |
daily backups (--backups=false turns them off) |
--deletion-protection |
boolean | refuse deletes of the warehouse and its databases until it’s turned off | |
--description |
string | what it’s for | |
--engine-version |
string | the ClickHouse version, from ezgh warehouse warehouses engine-versions | |
--from-backup |
string | restore this backup, <whs_ ID>/<backup ID> | |
--instance-type |
string | every node’s instance type, with at least 8 GiB of memory | |
--keeper |
string | the Keeper mode: dedicated or colocated (default dedicated for more than one node) | |
--maintenance-window |
string | when maintenance happens, ddd:HH:MM-ddd:HH:MM UTC, at least 30 minutes | |
--plaintext-port |
boolean | also serve HTTP on 8123 and the native protocol on 9000 without TLS, inside the VPC | |
--replicas |
int | how many replicas of each shard, 1 to 3 (default 1) | |
--shards |
int | how many shards, 1 to 8 (default 1) | |
--storage-size-gib |
int | each node’s storage, in GiB (at least 10) | |
--subnet |
string | the subnet it’s in (sbn_…) | |
--tag |
string, repeatable | a tag, key=value; repeat for more | |
--wait |
boolean | wait until it’s available, and print it then | |
--wait-timeout |
duration | 10m0s |
how long --wait waits |
This command also takes the global flags.