Description
It looks like in provisioned mode, ns_server starts cbq-engine with a flag "sequential_scan_disabled" set to true. [ MB-60132 ]
But query has not defined this "sequential_scan_disabled" flag. This causes the query engine to not start.
Repro steps:
1. Start the cluster ( with the Query service ) in provisioned mode.
a. Cluster_run:
./cluster_run --provisioned
|
Or
b. Couchbase Server Installation:
mkdir -p /etc/couchbase.d && echo 'provisioned' > /etc/couchbase.d/config_profile |
Start couchbase server.
2. The query service does not start.
In query.log:
2023-12-20T12:28:16.979+05:30 [INFO] cbq-engine starting version=7.6.0-N1QL go-version=go1.21.3 pid=26268 |
2023-12-20T12:28:16.979+05:30 [INFO] Current nofiles rlimit: 4096 (max: 4096) |
flag provided but not defined: -sequential_scan_disabled
|
Usage of /Users/dhanyagowrish/elixir/trinity-test/install/bin/cbq-engine:
|
-acctstore string
|
Accounting store address (http://URL or stub:) (default "gometrics:") |
-auto-prepare
|
Silently prepare ad hoc statements if possible |
-cafile string
|
HTTPS CA certificates
|
-certfile string
|
HTTPS certificate chain file
|
-clientCertFile string
|
Internal communications certificate
|
-clientKeyFile string
|
Internal communications private key |
-completed-limit int |
maximum number of completed requests (default 4000) |
-completed-threshold int |
cache completed query lasting longer than this many milliseconds (default -1) |
-configstore string
|
Configuration store address (http://URL or stub:) (default "stub:") |
-controls
|
Response to include controls section
|
-cpuprofile string
|
write cpu profile to file
|
-datastore string
|
Datastore address (http://URL or dir:PATH or mock:) |
-debug
|
Debug mode
|
-deploymentModel string
|
Deployment Model: default, serverless, provisioned (default "default") |
-dictionary-cache-limit int |
maximum number of entries in dictionary cache (default 16384) |
-enterprise
|
Enterprise mode (default true) |
-functions-limit int |
maximum number of cached functions (default 16384) |
-gc-percent int |
Go runtime garbage collection target percentage (default 200) |
-http string
|
HTTP service address (default ":8093") |
-https string
|
HTTPS service address (default ":18093") |
-ipv4 string
|
Query uses IPv4 listeners only (default "required") |
-ipv6 string
|
Query is IPv6 compliant (default "optional") |
-keep-alive-length int |
maximum size of buffered result (default 16384) |
-keyfile string
|
HTTPS private key file |
-logger string
|
Logger implementation
|
-loglevel string
|
Log level: debug, trace, info, warn, error, severe, none (default "info") |
-max-index-api int |
Max Index API (default 4) |
-max-parallelism int |
Maximum parallelism per query; use zero or negative value to use maximum (default 1) |
-memory-quota uint
|
Maximum amount of document memory allowed per request, in MB
|
-memprofile string
|
write memory profile to this file |
-metrics
|
Whether to provide metrics (default true) |
-mutation-limit int |
Maximum LIMIT for data modification statements; use zero or negative value to disable |
-n1ql-feat-ctrl uint
|
N1QL Feature Controls (default 76) |
-namespace string
|
Default namespace (default "default") |
-node-quota uint
|
Soft memory limit per node, in MB
|
-node-quota-val-percent uint
|
Percentage of node quota reserved for value memory (0-100) (default 67) |
-order-limit int |
Maximum LIMIT for ORDER BY clauses; use zero or negative value to disable |
-pipeline-batch int |
Number of items execution operators can batch (default 16) |
-pipeline-cap int |
Maximum number of items each execution operator can buffer (default 512) |
-plus-servicers int |
Plus servicer count
|
-prepared-limit int |
maximum number of prepared statements (default 16384) |
-pretty
|
Pretty output
|
-profile string
|
Profiling state: off, phases, timings (default "off") |
-readonly
|
Read-only mode
|
-regulatorSettingsFile string
|
Regulator settings file
|
-request-cap int |
Maximum number of queued requests per logical CPU (default 256) |
-request-error-limit int |
Maximum number of errors to accumulate before aborting a request (default 16) |
-request-size-cap int |
Maximum size of a request (default 67108864) |
-scan-cap int |
Maximum buffer size for index scans; use zero or negative value to disable (default 512) |
-servicers int |
Servicer count
|
-signature
|
Whether to provide signature (default true) |
-static-path string |
Path to static content (default "static") |
-tasks-limit int |
maximum number of cached tasks (default 16384) |
-timeout duration
|
Server execution timeout, e.g. 500ms or 2s; use zero or negative value to disable
|
-txtimeout duration
|
Maximum Transaction timeout, e.g. 2m or 2s; use zero or negative to use request level value
|
-use-replica string
|
Allow reading from replica vBuckets (default "unset") |
-uuid string
|
Node UUID.
|
cc Kamini Jagtiani , Donald Haggart
Attachments
Issue Links
- is caused by
-
MB-60132 sequential scan should be disabled in Capella provisioned profile
- Closed