aboutsummaryrefslogtreecommitdiffstats
path: root/docs/model
diff options
context:
space:
mode:
authorVratko Polak <vrpolak@cisco.com>2021-12-15 17:14:36 +0100
committerVratko Polak <vrpolak@cisco.com>2021-12-15 17:14:36 +0100
commit01d8f262afc567c3d49a23c3cb2cdeaced8a6887 (patch)
tree0449c972d8201be16d648dd749e0a7d116aa8b71 /docs/model
parentcca05a55f3434d8a031b98f4a496adb8df20c122 (diff)
UTI: Export results
+ Model version 1.0.0. - Only some result types are exported. + MRR, NDRPDR and SOAK. - Other result types to be added later. + In contrast, all test types are detected. + Convert custom classes to JSON-serializable equivalents. + Sort dict keys before converting to JSON. + Override the order for some known keys. + Export sets as sorted arrays. + Convert to info content from serialized raw content. + Also export outputs for suite setups and teardowns. + Info files for setup/teardown exist only temporarily. + The data is merged into suite.info.json file. + This simplifies presentation of total suite duration. + Define model via JSON schema: - Just test case, suite setup/teardown/suite to be added later. - Just info, raw to be added later. + Proper descriptions. + Json is generated from yaml. + This is a convenience for maintainers. + The officially used schema is the .json one. + TODOs written into a separate .txt file. + Validate exported instance against the schema. + Include format checking. + Update CSIT requirements for validation dependencies. + This needs python-dateutil==2.8.2, only a patch bump. + Compute bandwidth also for soak tests. + This unifies with NDRPDR to simplify schema definition. - PAL may need an update for parsing soak test message. + Include SSH log items, raw output only. + Generate all outputs in a single filesystem tree. + Move raw outputs into test_output_raw.tar.xz. + Rename existing tar with suites to generated_robot_files.tar.xz. Change-Id: I69ff7b330ed1a14dc435fd0ef008e753c0d7f78c Signed-off-by: Vratko Polak <vrpolak@cisco.com>
Diffstat (limited to 'docs/model')
-rw-r--r--docs/model/current/schema/test_case.info.schema.json571
-rw-r--r--docs/model/current/schema/test_case.info.schema.yaml637
-rw-r--r--docs/model/current/schema/todos.txt16
-rw-r--r--docs/model/current/schema/yaml2json.py29
-rw-r--r--docs/model/current/top.rst43
5 files changed, 1292 insertions, 4 deletions
diff --git a/docs/model/current/schema/test_case.info.schema.json b/docs/model/current/schema/test_case.info.schema.json
new file mode 100644
index 0000000000..358a3e7582
--- /dev/null
+++ b/docs/model/current/schema/test_case.info.schema.json
@@ -0,0 +1,571 @@
+{
+ "$id": "https://fd.io/FIXME/CSIT/UTI/test_case/info/0.2.0",
+ "$schema": "https://json-schema.org/draft/2020-12/schema",
+ "description": "Schema for info output of test case.",
+ "allOf": [
+ {
+ "description": "The main structure, without conditional relations between fields yet.",
+ "type": "object",
+ "additionalProperties": false,
+ "properties": {
+ "duration": {
+ "description": "A derived quantity, present only in info output files. Difference between start_time and end_time, in seconds.",
+ "$ref": "#/$defs/types/nonnegative_number"
+ },
+ "dut_type": {
+ "description": "DUT type used, e.g. VPP or DPDK.",
+ "type": "string"
+ },
+ "dut_version": {
+ "description": "Version string appropriate to DUT type used.",
+ "type": "string"
+ },
+ "end_time": {
+ "description": "UTC date and time in RFC 3339 format, specifying calendar time just before test case ended (at the end of test case teardown).",
+ "$ref": "#/$defs/types/date_time"
+ },
+ "hosts": {
+ "description": "Array of hosts this test interacted with. This can be used for identifying testbed number. Valid tests shoud interact with at least one DUT or TG. The array is usually sorted, but that is not a requirement.",
+ "type": "array",
+ "minItems": 1,
+ "items": {
+ "description": "Host identifier, usually numeric IPv4 address.",
+ "type": "string"
+ }
+ },
+ "log": {
+ "description": "No log items are implemented in the current version, but the (empty) list is present to simplify logic in multi-version importers.",
+ "$ref": "#/$defs/types/empty_array"
+ },
+ "message": {
+ "description": "If passed is true, this value is empty. Otherwise, value taken directly from TEST_MESSAGE Robot variable, read at the end of test case (in test teardown, before export and validation). It contains information from the exception that caused the failure, probably with additional exceptions from teardown keywords.",
+ "type": "string"
+ },
+ "passed": {
+ "description": "Value set accordingly to TEST_STATUS Robot variable, true if and only if the status is \"PASS\". The status is read at the end of test case (in test teardown, before export and validation).",
+ "type": "boolean"
+ },
+ "result": {
+ "type": "object",
+ "allOf": [
+ {
+ "description": "Sub-schema common for all cases, only result type identifier defined here.",
+ "properties": {
+ "type": {
+ "description": "Identifier of which result type case is applied.",
+ "type": "string"
+ }
+ },
+ "required": [
+ "type"
+ ]
+ },
+ {
+ "oneOf": [
+ {
+ "description": "Result type for unknown case. This case represents a test with no specific result (outside message), e.g. device test; or a test with result not parsed into this version of model yet, e.g. GSO test.",
+ "additionalProperties": false,
+ "properties": {
+ "type": {
+ "const": "unknown"
+ }
+ }
+ },
+ {
+ "description": "Result type MRR case.",
+ "additionalProperties": false,
+ "properties": {
+ "type": {
+ "const": "mrr"
+ },
+ "receive_rate": {
+ "description": "The results refer to receive rates for multiple MRR trials. For PPS, these are aggregate (bidirectional) rates. Currently, the tests are exporting approximated receive rates. That means the actual trial duration is measured (as opposed to trusting traffic generator to honor its target duration), so the resulting values contain noise from time measurement, and can be lower than the real performance (due to various time overheads). Bandwidth values are supported, but currently Robot does not export them.",
+ "$ref": "#/$defs/types/rate_list_with_bandwidth"
+ }
+ },
+ "required": [
+ "type",
+ "receive_rate"
+ ]
+ },
+ {
+ "description": "Result type NDRPDR case.",
+ "additionalProperties": false,
+ "properties": {
+ "type": {
+ "const": "ndrpdr"
+ },
+ "ndr": {
+ "description": "The results refer to search for NDR (Non Drop Rate). For PPS, this is aggregate (bidirectional) rate. Each bound was used as the target load value in a full-duration trial measurement. The accepted loss ratio for NDR is exact zero. Note that packets the Traffic Generator did not send are also counted as lost packets.",
+ "$ref": "#/$defs/macros/lower_and_upper_rate"
+ },
+ "pdr": {
+ "description": "The results refer to search for PDR (Partial Drop Rate). For PPS, this is aggregate (bidirectional) rate. Each bound was used as the target load value in a full-duration trial measurement. The accepted loss ratio for PDR is 0.5%. Note that packets the Traffic Generator did not send are also counted as lost packets.",
+ "$ref": "#/$defs/macros/lower_and_upper_rate"
+ },
+ "latency_forward": {
+ "description": "Object with results related to latency part of NDRPDR test, for forward traffic direction. It is the direction used in unidirectional traffic profiles. ASTF profiles and IMIX STL profiles do not support latency information, so for those tests this object is missing. It is also missing if Traffic Generator fails to return valid latency results for any other reasons, e.g. latency rate is too high for CPU/NIC used.",
+ "$ref": "#/$defs/macros/latency_for_loads"
+ },
+ "latency_reverse": {
+ "description": "Object with results related to latency part of NDRPDR test, for reverse traffic diration. This object is not present when unidirectional traffic profiles are used. ASTF profiles and IMIX STL profiles do not support latency information, so for those tests this object is missing. It is also missing if Traffic Generator fails to return valid latency results for any other reasons, e.g. latency rate is too high for CPU/NIC used.",
+ "$ref": "#/$defs/macros/latency_for_loads"
+ }
+ },
+ "required": [
+ "type",
+ "ndr",
+ "pdr"
+ ]
+ },
+ {
+ "description": "Result type SOAK case.",
+ "additionalProperties": false,
+ "properties": {
+ "type": {
+ "const": "soak"
+ },
+ "critical_rate": {
+ "description": "The results refer to bayesian estimate of critical rate corresponding to average loss ratio of 10^-7. For PPS, this is aggregate (bidirectional) rate. The bounds are computed from trial measurement results, but are not equal to any target load used. Note that packets the Traffic Generator did not send are also counted as lost packets.",
+ "$ref": "#/$defs/macros/lower_and_upper_rate"
+ }
+ },
+ "required": [
+ "type",
+ "critical_rate"
+ ]
+ }
+ ]
+ }
+ ]
+ },
+ "start_time": {
+ "description": "UTC date and time in RFC 3339 format, specifying calendar time just after test case started (at the start of test setup).",
+ "$ref": "#/$defs/types/date_time"
+ },
+ "tags": {
+ "description": "The list of strings comes directly from Robot variable TEST_TAGS. The content should include both static and dynamic tags at the end of test case (teardown).",
+ "type": "array",
+ "items": {
+ "type": "string"
+ }
+ },
+ "test_documentation": {
+ "description": "Value taken directly from TEST_DOCUMENTATION Robot variable. The content is what you see in suite file at test case definition, which is usually empty as CSIT uses data driven test cases.",
+ "type": "string"
+ },
+ "test_id": {
+ "description": "A derived quantity, present only in info output files. It is the most complete and unique identifier for a test case. This property has a value, of the following form: {suite_name}.{test_name} Here, suite name comes from SUITE_NAME robot variable, test name comes from TEST_NAME robot variable, but both are converted to lower case, and spaces are replaced by underscores.",
+ "type": "string",
+ "minLength": 3
+ },
+ "test_name_long": {
+ "description": "A derived quantity, present only in info output files. This property has a value, of the following form: {nic_short_name}-{frame_size}-{threads_and_cores}-{suite_part} Here, suite part is very similar to suite tag, but additionally may contain a prefix describing NIC driver used (if it is not the default one, drv_vfio_pci for VPP tests). Any space is replaced by underscore and letters are lower case.",
+ "type": "string",
+ "minLength": 3
+ },
+ "test_name_short": {
+ "description": "A derived quantity, present only in info output files. This property has a value very similar to suite tag, but additionally may contain a prefix describing NIC driver used (if it is not the default one, drv_vfio_pci for VPP tests). Any space is replaced by underscore and letters are lower case.",
+ "type": "string",
+ "minLength": 3
+ },
+ "test_type": {
+ "description": "A derived quantity, present only in info output files. Test type identifier, PAL uses it to group similar tests, e.g. for comparison tables. Ideally, this information should be parseable from test name, but the current naming scheme is not simple/consistent enough. The current implementation queries the robot test tags. The resulting value is frequently identical to result type, but this schema version does not require any relation there, as PAL may want to group tests differently.",
+ "type": "string",
+ "enum": [
+ "device",
+ "gso",
+ "hoststack",
+ "mrr",
+ "ndrpdr",
+ "reconf",
+ "soak",
+ "vsap"
+ ]
+ },
+ "version": {
+ "description": "CSIT model version (semver format) the exporting code adhered to.",
+ "type": "string",
+ "const": "1.0.0"
+ }
+ },
+ "required": [
+ "duration",
+ "dut_type",
+ "dut_version",
+ "end_time",
+ "hosts",
+ "log",
+ "message",
+ "passed",
+ "result",
+ "start_time",
+ "tags",
+ "test_documentation",
+ "test_id",
+ "test_name_long",
+ "test_name_short",
+ "test_type",
+ "version"
+ ]
+ },
+ {
+ "description": "Subschema validating relation between status and message.",
+ "oneOf": [
+ {
+ "description": "Subschema for passing tests, message has to be empty.",
+ "type": "object",
+ "properties": {
+ "passed": {
+ "const": true
+ },
+ "message": {
+ "const": ""
+ }
+ }
+ },
+ {
+ "description": "Subschema for failing tests, mesage cannot be empty.",
+ "type": "object",
+ "properties": {
+ "passed": {
+ "const": false
+ },
+ "message": {
+ "minLength": 1
+ }
+ }
+ }
+ ]
+ },
+ {
+ "description": "Subschema validating relation between dut_type and dut_version.",
+ "oneOf": [
+ {
+ "description": "Subschema for tests with no DUT, e.g. TRex self-test.",
+ "type": "object",
+ "properties": {
+ "dut_type": {
+ "const": "none"
+ },
+ "dut_version": {
+ "const": ""
+ }
+ }
+ },
+ {
+ "description": "Subschema for DUT type VPP.",
+ "type": "object",
+ "properties": {
+ "dut_type": {
+ "const": "VPP"
+ },
+ "dut_version": {
+ "minLength": 1
+ }
+ }
+ },
+ {
+ "description": "Subschema for DUT type DPDK.",
+ "type": "object",
+ "properties": {
+ "dut_type": {
+ "const": "DPDK"
+ },
+ "dut_version": {
+ "minLength": 1
+ }
+ }
+ }
+ ]
+ }
+ ],
+ "$defs": {
+ "types": {
+ "nonnegative_number": {
+ "type": "number",
+ "minimum": 0
+ },
+ "positive_number": {
+ "type": "number",
+ "minimum": 1
+ },
+ "nonnegative_integer": {
+ "type": "integer",
+ "minimum": 0
+ },
+ "positive_integer": {
+ "type": "integer",
+ "minimum": 1
+ },
+ "date_time": {
+ "type": "string",
+ "format": "date-time"
+ },
+ "empty_array": {
+ "type": "array",
+ "maxItems": 0
+ },
+ "rate_unit": {
+ "description": "Packets per second (pps) or connections per second (cps).",
+ "type": "string",
+ "enum": [
+ "pps",
+ "cps"
+ ]
+ },
+ "bandwidth_unit": {
+ "description": "Unit of measurement for bandwidth values. Currently a constant, but later versions of model may allow more units.",
+ "enum": [
+ "bps"
+ ]
+ },
+ "value_with_unit": {
+ "description": "Reusable composite type, value together with its unit of measurement.",
+ "type": "object",
+ "additionalProperties": false,
+ "properties": {
+ "value": {
+ "description": "Numeric value, context specified elsewhere. The only assumption is that value is not negative.",
+ "$ref": "#/$defs/types/nonnegative_number"
+ },
+ "unit": {
+ "description": "Unit of measurement for the value. Context and allowed values are specified elsewhere.",
+ "type": "string"
+ }
+ },
+ "required": [
+ "value",
+ "unit"
+ ]
+ },
+ "rate_without_bandwidth": {
+ "description": "Reusable type, for various rate quantites.",
+ "allOf": [
+ {
+ "$ref": "#/$defs/types/value_with_unit"
+ },
+ {
+ "properties": {
+ "value": {
+ "description": "Unless specified otherwise, this is the aggregated rate (sum of both traffic directions). Depending on the usage, the value can express intended load, offered load, receive rate, and various approximations or estimated bounds thereof."
+ },
+ "unit": {
+ "description": "A transaction rate unit the value is expressed in.",
+ "$ref": "#/$defs/types/rate_unit"
+ }
+ }
+ }
+ ]
+ },
+ "bandwidth": {
+ "description": "Reusable type, for various bandwidth quantites.",
+ "allOf": [
+ {
+ "$ref": "#/$defs/types/value_with_unit"
+ },
+ {
+ "properties": {
+ "value": {
+ "description": "Bandwidth value computed from the corresponding rate."
+ },
+ "unit": {
+ "$ref": "#/$defs/types/bandwidth_unit"
+ }
+ }
+ }
+ ]
+ },
+ "rate_with_bandwidth": {
+ "description": "Reusable composite type, joining primary rate with optional derived bandwidth. Not all test types currently compute bandwidth, even if rate unit is pps.",
+ "type": "object",
+ "additionalProperties": false,
+ "properties": {
+ "rate": {
+ "$ref": "#/$defs/types/rate_without_bandwidth"
+ },
+ "bandwidth": {
+ "$ref": "#/$defs/types/bandwidth"
+ }
+ },
+ "required": [
+ "rate"
+ ]
+ },
+ "value_list_with_unit_and_stats": {
+ "description": "Reusable composite type, multiple values together with their unit of measurement and derived statistics.",
+ "type": "object",
+ "additionalProperties": false,
+ "properties": {
+ "values": {
+ "description": "List of values of the same unit, useful for MRR.",
+ "type": "array",
+ "minItmes": 1,
+ "items": {
+ "description": "Numeric value, context specified elsewhere. The only assumption is that the value is nonnegative.",
+ "$ref": "#/$defs/types/nonnegative_number"
+ }
+ },
+ "avg": {
+ "description": "A derived quantity, present only in info output files. It is the arithmetic average of the values list.",
+ "$ref": "#/$defs/types/nonnegative_number"
+ },
+ "stdev": {
+ "description": "A derived quantity, present only in info output files. It is the standard deviation for the values list, as computed by jumpavg library.",
+ "$ref": "#/$defs/types/nonnegative_number"
+ },
+ "unit": {
+ "description": "Unit of measurement for the values. Context and allowed values are specified elsewhere.",
+ "type": "string"
+ }
+ },
+ "required": [
+ "values",
+ "avg",
+ "stdev",
+ "unit"
+ ]
+ },
+ "rate_list_without_bandwidth": {
+ "description": "Reusable composite type, multiple rate values.",
+ "allOf": [
+ {
+ "$ref": "#/$defs/types/value_list_with_unit_and_stats"
+ },
+ {
+ "properties": {
+ "values": {
+ "items": {
+ "description": "Unless specified otherwise, this is the aggregated rate (sum of both traffic directions). Depending on the usage, the value can express intended load, offered load, receive rate, and various approximations or estimated bounds thereof."
+ }
+ },
+ "unit": {
+ "$ref": "#/$defs/types/rate_unit"
+ }
+ }
+ }
+ ]
+ },
+ "bandwidth_list": {
+ "description": "Reusable composite type, multiple bandwidth values. This is a derived entity, thus it only appears in info output, and only if rate unit is pps.",
+ "allOf": [
+ {
+ "$ref": "#/$defs/types/value_list_with_unit_and_stats"
+ },
+ {
+ "properties": {
+ "values": {
+ "items": {
+ "description": "Unless specified otherwise, this is the aggregated bandwidth (sum of both traffic directions). Depending on the usage, the value can express intended load, offered load, receive rate, and various approximations or estimated bounds thereof."
+ }
+ },
+ "unit": {
+ "$ref": "#/$defs/types/bandwidth_unit"
+ }
+ }
+ }
+ ]
+ },
+ "rate_list_with_bandwidth": {
+ "description": "Reusable composite type, joining primary rates with optional derived bandwidths (and stats). No test types currently computes the bandwidth part.",
+ "type": "object",
+ "additionalProperties": false,
+ "properties": {
+ "rate": {
+ "$ref": "#/$defs/types/rate_list_without_bandwidth"
+ },
+ "bandwidth": {
+ "$ref": "#/$defs/types/bandwidth_list"
+ }
+ },
+ "required": [
+ "rate"
+ ]
+ }
+ },
+ "macros": {
+ "lower_and_upper_rate": {
+ "type": "object",
+ "additionalProperties": false,
+ "properties": {
+ "lower": {
+ "description": "The lower bound (or min_rate) for the estimate of a particular searched value.",
+ "$ref": "#/$defs/types/rate_with_bandwidth"
+ },
+ "upper": {
+ "description": "The upper bound (or max_rate) for the estimate of a particular searched value.",
+ "$ref": "#/$defs/types/rate_with_bandwidth"
+ }
+ },
+ "required": [
+ "lower",
+ "upper"
+ ]
+ },
+ "latency_numbers": {
+ "type": "object",
+ "additionalProperties": false,
+ "properties": {
+ "min": {
+ "description": "Rounded minimal latency time measured in this trial. See unit property for the unit of measurement.",
+ "$ref": "#/$defs/types/nonnegative_integer"
+ },
+ "max": {
+ "description": "Rounded maximal latency time measured in this trial. See unit property for the unit of measurement. Zero value is not allowed, as that is one of symptoms of Traffic Generator failing to get proper latency.",
+ "$ref": "#/$defs/types/positive_integer"
+ },
+ "avg": {
+ "description": "Rounded average latency time measured in this trial. See unit property for the unit of measurement.",
+ "$ref": "#/$defs/types/nonnegative_integer"
+ },
+ "hdrh": {
+ "description": "Base64-encoded compressed representation of HDRHistogram of all latency sample times encountered in this latency trial. See unit property for the unit of measurement. Note that some bins can be several units wide.",
+ "type": "string"
+ },
+ "unit": {
+ "description": "Unit of measurement for latency times. Currently a constant, but later versions of the model may allow more values.",
+ "type": "string",
+ "enum": [
+ "us"
+ ]
+ }
+ },
+ "required": [
+ "avg",
+ "hdrh",
+ "max",
+ "min",
+ "unit"
+ ]
+ },
+ "latency_for_loads": {
+ "type": "object",
+ "additionalProperties": false,
+ "properties": {
+ "pdr_0": {
+ "description": "Object related to latency measurement performed at minimal rate (currently 9000 pps per direction).",
+ "$ref": "#/$defs/macros/latency_numbers"
+ },
+ "pdr_10": {
+ "description": "Object related to latency measurement performed at 10% of PDR lower bound, if needed rounded up to minimal rate (currently 9000 pps per direction).",
+ "$ref": "#/$defs/macros/latency_numbers"
+ },
+ "pdr_50": {
+ "description": "Object related to latency measurement performed at 50% of PDR lower bound, if needed rounded up to minimal rate (currently 9000 pps per direction).",
+ "$ref": "#/$defs/macros/latency_numbers"
+ },
+ "pdr_90": {
+ "description": "Object related to latency measurement performed at 90% of PDR lower bound, if needed rounded up to minimal rate (currently 9000 pps per direction).",
+ "$ref": "#/$defs/macros/latency_numbers"
+ }
+ },
+ "required": [
+ "pdr_0",
+ "pdr_10",
+ "pdr_50",
+ "pdr_90"
+ ]
+ }
+ }
+ }
+} \ No newline at end of file
diff --git a/docs/model/current/schema/test_case.info.schema.yaml b/docs/model/current/schema/test_case.info.schema.yaml
new file mode 100644
index 0000000000..7771d7afec
--- /dev/null
+++ b/docs/model/current/schema/test_case.info.schema.yaml
@@ -0,0 +1,637 @@
+# Copyright (c) 2021 Cisco and/or its affiliates.
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at:
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+---
+
+$id: https://fd.io/FIXME/CSIT/UTI/test_case/info/0.2.0
+$schema: https://json-schema.org/draft/2020-12/schema
+description: >-
+ Schema for info output of test case.
+allOf:
+- description: >-
+ The main structure, without conditional relations between fields yet.
+ type: object
+ additionalProperties: false
+ properties:
+ duration:
+ description: >-
+ A derived quantity, present only in info output files.
+ Difference between start_time and end_time, in seconds.
+ $ref: "#/$defs/types/nonnegative_number"
+ dut_type:
+ description: >-
+ DUT type used, e.g. VPP or DPDK.
+ type: string
+ dut_version:
+ description: >-
+ Version string appropriate to DUT type used.
+ type: string
+ end_time:
+ description: >-
+ UTC date and time in RFC 3339 format, specifying calendar time
+ just before test case ended (at the end of test case teardown).
+ $ref: "#/$defs/types/date_time"
+ hosts:
+ description: >-
+ Array of hosts this test interacted with.
+ This can be used for identifying testbed number.
+ Valid tests shoud interact with at least one DUT or TG.
+ The array is usually sorted, but that is not a requirement.
+ type: array
+ minItems: 1
+ items:
+ description: >-
+ Host identifier, usually numeric IPv4 address.
+ type: string
+ log:
+ description: >-
+ No log items are implemented in the current version,
+ but the (empty) list is present to simplify logic
+ in multi-version importers.
+ $ref: "#/$defs/types/empty_array"
+ message:
+ description: >-
+ If passed is true, this value is empty.
+ Otherwise, value taken directly from TEST_MESSAGE
+ Robot variable, read at the end of test case
+ (in test teardown, before export and validation).
+ It contains information from the exception
+ that caused the failure, probably with additional
+ exceptions from teardown keywords.
+ type: string
+ passed:
+ description: >-
+ Value set accordingly to TEST_STATUS Robot variable,
+ true if and only if the status is "PASS".
+ The status is read at the end of test case
+ (in test teardown, before export and validation).
+ type: boolean
+ result:
+ type: object
+ allOf:
+ - description: >-
+ Sub-schema common for all cases,
+ only result type identifier defined here.
+ properties:
+ type:
+ description: >-
+ Identifier of which result type case is applied.
+ type: string
+ required:
+ - type
+ - oneOf:
+ - description: >-
+ Result type for unknown case.
+ This case represents a test with no specific result
+ (outside message), e.g. device test;
+ or a test with result not parsed into
+ this version of model yet, e.g. GSO test.
+ additionalProperties: false
+ properties:
+ type:
+ const: unknown
+ - description: >-
+ Result type MRR case.
+ additionalProperties: false
+ properties:
+ type:
+ const: mrr
+ receive_rate:
+ description: >-
+ The results refer to receive rates for multiple
+ MRR trials. For PPS, these are aggregate
+ (bidirectional) rates.
+ Currently, the tests are exporting
+ approximated receive rates.
+ That means the actual trial duration
+ is measured (as opposed to trusting traffic
+ generator to honor its target duration),
+ so the resulting values contain noise
+ from time measurement, and can be lower
+ than the real performance
+ (due to various time overheads).
+ Bandwidth values are supported, but currently
+ Robot does not export them.
+ $ref: "#/$defs/types/rate_list_with_bandwidth"
+ required:
+ - type
+ - receive_rate
+ - description: >-
+ Result type NDRPDR case.
+ additionalProperties: false
+ properties:
+ type:
+ const: ndrpdr
+ ndr:
+ description: >-
+ The results refer to search for NDR
+ (Non Drop Rate). For PPS, this is aggregate
+ (bidirectional) rate.
+ Each bound was used as the target load value
+ in a full-duration trial measurement.
+ The accepted loss ratio for NDR is exact zero.
+ Note that packets the Traffic Generator
+ did not send are also counted as lost packets.
+ $ref: "#/$defs/macros/lower_and_upper_rate"
+ pdr:
+ description: >-
+ The results refer to search for PDR
+ (Partial Drop Rate). For PPS, this is aggregate
+ (bidirectional) rate.
+ Each bound was used as the target load value
+ in a full-duration trial measurement.
+ The accepted loss ratio for PDR is 0.5%.
+ Note that packets the Traffic Generator
+ did not send are also counted as lost packets.
+ $ref: "#/$defs/macros/lower_and_upper_rate"
+ latency_forward:
+ description: >-
+ Object with results related to latency part
+ of NDRPDR test, for forward traffic direction.
+ It is the direction used in unidirectional
+ traffic profiles.
+ ASTF profiles and IMIX STL profiles
+ do not support latency information,
+ so for those tests this object is missing.
+ It is also missing if Traffic Generator
+ fails to return valid latency results
+ for any other reasons,
+ e.g. latency rate is too high for CPU/NIC used.
+ $ref: "#/$defs/macros/latency_for_loads"
+ latency_reverse:
+ description: >-
+ Object with results related to latency part
+ of NDRPDR test, for reverse traffic diration.
+ This object is not present
+ when unidirectional traffic profiles are used.
+ ASTF profiles and IMIX STL profiles
+ do not support latency information,
+ so for those tests this object is missing.
+ It is also missing if Traffic Generator
+ fails to return valid latency results
+ for any other reasons,
+ e.g. latency rate is too high for CPU/NIC used.
+ $ref: "#/$defs/macros/latency_for_loads"
+ required:
+ - type
+ - ndr
+ - pdr
+ - description: >-
+ Result type SOAK case.
+ additionalProperties: false
+ properties:
+ type:
+ const: soak
+ critical_rate:
+ description: >-
+ The results refer to bayesian estimate
+ of critical rate corresponding to
+ average loss ratio of 10^-7.
+ For PPS, this is aggregate (bidirectional) rate.
+ The bounds are computed from
+ trial measurement results,
+ but are not equal to any target load used.
+ Note that packets the Traffic Generator
+ did not send are also counted as lost packets.
+ $ref: "#/$defs/macros/lower_and_upper_rate"
+ required:
+ - type
+ - critical_rate
+ start_time:
+ description: >-
+ UTC date and time in RFC 3339 format, specifying calendar time
+ just after test case started (at the start of test setup).
+ $ref: "#/$defs/types/date_time"
+ tags:
+ description: >-
+ The list of strings comes directly
+ from Robot variable TEST_TAGS.
+ The content should include both static and dynamic tags
+ at the end of test case (teardown).
+ type: array
+ items:
+ type: string
+ test_documentation:
+ description: >-
+ Value taken directly from TEST_DOCUMENTATION Robot variable.
+ The content is what you see in suite file
+ at test case definition, which is usually empty
+ as CSIT uses data driven test cases.
+ type: string
+ test_id:
+ description: >-
+ A derived quantity, present only in info output files.
+ It is the most complete and unique identifier for a test case.
+ This property has a value, of the following form:
+ {suite_name}.{test_name}
+ Here, suite name comes from SUITE_NAME robot variable,
+ test name comes from TEST_NAME robot variable,
+ but both are converted to lower case,
+ and spaces are replaced by underscores.
+ type: string
+ minLength: 3
+ test_name_long:
+ description: >-
+ A derived quantity, present only in info output files.
+ This property has a value, of the following form:
+ {nic_short_name}-{frame_size}-{threads_and_cores}-{suite_part}
+ Here, suite part is very similar to suite tag,
+ but additionally may contain a prefix describing NIC driver used
+ (if it is not the default one, drv_vfio_pci for VPP tests).
+ Any space is replaced by underscore and letters are lower case.
+ type: string
+ minLength: 3
+ test_name_short:
+ description: >-
+ A derived quantity, present only in info output files.
+ This property has a value very similar to suite tag,
+ but additionally may contain a prefix describing NIC driver used
+ (if it is not the default one, drv_vfio_pci for VPP tests).
+ Any space is replaced by underscore and letters are lower case.
+ type: string
+ minLength: 3
+ test_type:
+ description: >-
+ A derived quantity, present only in info output files.
+ Test type identifier, PAL uses it to group similar tests,
+ e.g. for comparison tables.
+ Ideally, this information should be parseable from test name,
+ but the current naming scheme is not simple/consistent enough.
+ The current implementation queries the robot test tags.
+ The resulting value is frequently identical to result type,
+ but this schema version does not require any relation there,
+ as PAL may want to group tests differently.
+ type: string
+ enum:
+ - device
+ - gso
+ - hoststack
+ - mrr
+ - ndrpdr
+ - reconf
+ - soak
+ - vsap
+ version:
+ description: >-
+ CSIT model version (semver format)
+ the exporting code adhered to.
+ type: string
+ const: 1.0.0
+ required:
+ - duration
+ - dut_type
+ - dut_version
+ - end_time
+ - hosts
+ - log
+ - message
+ - passed
+ - result
+ - start_time
+ - tags
+ - test_documentation
+ - test_id
+ - test_name_long
+ - test_name_short
+ - test_type
+ - version
+- description: >-
+ Subschema validating relation between status and message.
+ oneOf:
+ - description: >-
+ Subschema for passing tests, message has to be empty.
+ type: object
+ properties:
+ passed:
+ const: true
+ message:
+ const: ""
+ - description: >-
+ Subschema for failing tests, mesage cannot be empty.
+ type: object
+ properties:
+ passed:
+ const: false
+ message:
+ minLength: 1
+- description: >-
+ Subschema validating relation between dut_type and dut_version.
+ oneOf:
+ - description: >-
+ Subschema for tests with no DUT, e.g. TRex self-test.
+ type: object
+ properties:
+ dut_type:
+ const: none
+ dut_version:
+ const: ""
+ - description: >-
+ Subschema for DUT type VPP.
+ type: object
+ properties:
+ dut_type:
+ const: VPP
+ dut_version:
+ minLength: 1
+ - description: >-
+ Subschema for DUT type DPDK.
+ type: object
+ properties:
+ dut_type:
+ const: DPDK
+ dut_version:
+ minLength: 1
+
+$defs:
+ types:
+ nonnegative_number:
+ type: number
+ minimum: 0
+ positive_number:
+ type: number
+ minimum: 1
+ nonnegative_integer:
+ type: integer
+ minimum: 0
+ positive_integer:
+ type: integer
+ minimum: 1
+ date_time:
+ type: string
+ format: date-time
+ empty_array:
+ type: array
+ maxItems: 0
+ rate_unit:
+ description: >-
+ Packets per second (pps) or connections per second (cps).
+ type: string
+ enum:
+ - pps
+ - cps
+ bandwidth_unit:
+ description: >-
+ Unit of measurement for bandwidth values.
+ Currently a constant, but later versions of model
+ may allow more units.
+ enum:
+ - bps
+ value_with_unit:
+ description: >-
+ Reusable composite type, value together with its
+ unit of measurement.
+ type: object
+ additionalProperties: false
+ properties:
+ value:
+ description: >-
+ Numeric value, context specified elsewhere.
+ The only assumption is that value is not negative.
+ $ref: "#/$defs/types/nonnegative_number"
+ unit:
+ description: >-
+ Unit of measurement for the value.
+ Context and allowed values are specified elsewhere.
+ type: string
+ required:
+ - value
+ - unit
+ rate_without_bandwidth:
+ description: >-
+ Reusable type, for various rate quantites.
+ allOf:
+ - $ref: "#/$defs/types/value_with_unit"
+ - properties:
+ value:
+ description: >-
+ Unless specified otherwise,
+ this is the aggregated rate
+ (sum of both traffic directions).
+ Depending on the usage, the value can express
+ intended load, offered load, receive rate,
+ and various approximations
+ or estimated bounds thereof.
+ unit:
+ description: >-
+ A transaction rate unit the value is expressed in.
+ $ref: "#/$defs/types/rate_unit"
+ bandwidth:
+ description: >-
+ Reusable type, for various bandwidth quantites.
+ allOf:
+ - $ref: "#/$defs/types/value_with_unit"
+ - properties:
+ value:
+ description: >-
+ Bandwidth value computed
+ from the corresponding rate.
+ unit:
+ $ref: "#/$defs/types/bandwidth_unit"
+ rate_with_bandwidth:
+ description: >-
+ Reusable composite type, joining primary rate
+ with optional derived bandwidth.
+ Not all test types currently compute bandwidth,
+ even if rate unit is pps.
+ type: object
+ additionalProperties: false
+ properties:
+ rate:
+ $ref: "#/$defs/types/rate_without_bandwidth"
+ bandwidth:
+ $ref: "#/$defs/types/bandwidth"
+ required:
+ - rate
+ value_list_with_unit_and_stats:
+ description: >-
+ Reusable composite type, multiple values together with their
+ unit of measurement and derived statistics.
+ type: object
+ additionalProperties: false
+ properties:
+ values:
+ description: >-
+ List of values of the same unit, useful for MRR.
+ type: array
+ minItmes: 1
+ items:
+ description: >-
+ Numeric value, context specified elsewhere.
+ The only assumption is that the value is nonnegative.
+ $ref: "#/$defs/types/nonnegative_number"
+ avg:
+ description: >-
+ A derived quantity, present only in info output files.
+ It is the arithmetic average of the values list.
+ $ref: "#/$defs/types/nonnegative_number"
+ stdev:
+ description: >-
+ A derived quantity, present only in info output files.
+ It is the standard deviation for the values list,
+ as computed by jumpavg library.
+ $ref: "#/$defs/types/nonnegative_number"
+ unit:
+ description: >-
+ Unit of measurement for the values.
+ Context and allowed values are specified elsewhere.
+ type: string
+ required:
+ - values
+ - avg
+ - stdev
+ - unit
+ rate_list_without_bandwidth:
+ description: >-
+ Reusable composite type, multiple rate values.
+ allOf:
+ - $ref: "#/$defs/types/value_list_with_unit_and_stats"
+ - properties:
+ values:
+ items:
+ description: >-
+ Unless specified otherwise,
+ this is the aggregated rate
+ (sum of both traffic directions).
+ Depending on the usage, the value can express
+ intended load, offered load, receive rate,
+ and various approximations or estimated bounds
+ thereof.
+ unit:
+ $ref: "#/$defs/types/rate_unit"
+ bandwidth_list:
+ description: >-
+ Reusable composite type, multiple bandwidth values.
+ This is a derived entity, thus it only appears in info output,
+ and only if rate unit is pps.
+ allOf:
+ - $ref: "#/$defs/types/value_list_with_unit_and_stats"
+ - properties:
+ values:
+ items:
+ description: >-
+ Unless specified otherwise,
+ this is the aggregated bandwidth
+ (sum of both traffic directions).
+ Depending on the usage, the value can express
+ intended load, offered load, receive rate,
+ and various approximations or estimated bounds
+ thereof.
+ unit:
+ $ref: "#/$defs/types/bandwidth_unit"
+ rate_list_with_bandwidth:
+ description: >-
+ Reusable composite type, joining primary rates
+ with optional derived bandwidths (and stats).
+ No test types currently computes the bandwidth part.
+ type: object
+ additionalProperties: false
+ properties:
+ rate:
+ $ref: "#/$defs/types/rate_list_without_bandwidth"
+ bandwidth:
+ $ref: "#/$defs/types/bandwidth_list"
+ required:
+ - rate
+ macros:
+ lower_and_upper_rate:
+ type: object
+ additionalProperties: false
+ properties:
+ lower:
+ description: >-
+ The lower bound (or min_rate) for the estimate
+ of a particular searched value.
+ $ref: "#/$defs/types/rate_with_bandwidth"
+ upper:
+ description: >-
+ The upper bound (or max_rate) for the estimate
+ of a particular searched value.
+ $ref: "#/$defs/types/rate_with_bandwidth"
+ required:
+ - lower
+ - upper
+ latency_numbers:
+ type: object
+ additionalProperties: false
+ properties:
+ min:
+ description: >-
+ Rounded minimal latency time measured in this trial.
+ See unit property for the unit of measurement.
+ $ref: "#/$defs/types/nonnegative_integer"
+ max:
+ description: >-
+ Rounded maximal latency time measured in this trial.
+ See unit property for the unit of measurement.
+ Zero value is not allowed, as that is one of symptoms
+ of Traffic Generator failing to get proper latency.
+ $ref: "#/$defs/types/positive_integer"
+ avg:
+ description: >-
+ Rounded average latency time measured in this trial.
+ See unit property for the unit of measurement.
+ $ref: "#/$defs/types/nonnegative_integer"
+ hdrh:
+ description: >-
+ Base64-encoded compressed representation of HDRHistogram
+ of all latency sample times encountered
+ in this latency trial.
+ See unit property for the unit of measurement.
+ Note that some bins can be several units wide.
+ type: string
+ unit:
+ description: >-
+ Unit of measurement for latency times.
+ Currently a constant, but later versions
+ of the model may allow more values.
+ type: string
+ enum:
+ - us
+ required:
+ - avg
+ - hdrh
+ - max
+ - min
+ - unit
+ latency_for_loads:
+ type: object
+ additionalProperties: false
+ properties:
+ pdr_0:
+ description: >-
+ Object related to latency measurement performed
+ at minimal rate (currently 9000 pps per direction).
+ $ref: "#/$defs/macros/latency_numbers"
+ pdr_10:
+ description: >-
+ Object related to latency measurement performed
+ at 10% of PDR lower bound, if needed rounded up
+ to minimal rate (currently 9000 pps per direction).
+ $ref: "#/$defs/macros/latency_numbers"
+ pdr_50:
+ description: >-
+ Object related to latency measurement performed
+ at 50% of PDR lower bound, if needed rounded up
+ to minimal rate (currently 9000 pps per direction).
+ $ref: "#/$defs/macros/latency_numbers"
+ pdr_90:
+ description: >-
+ Object related to latency measurement performed
+ at 90% of PDR lower bound, if needed rounded up
+ to minimal rate (currently 9000 pps per direction).
+ $ref: "#/$defs/macros/latency_numbers"
+ required:
+ - pdr_0
+ - pdr_10
+ - pdr_50
+ - pdr_90
diff --git a/docs/model/current/schema/todos.txt b/docs/model/current/schema/todos.txt
new file mode 100644
index 0000000000..8e63b21a19
--- /dev/null
+++ b/docs/model/current/schema/todos.txt
@@ -0,0 +1,16 @@
+
+Export also tg_type and tg_version properties.
+
+Add description with link to methodology for MRR, NDRPDR and SOAK.
+
+Add multiplicity field to MRR result, so PAL can detect incomplete samples.
+
+Add link explaining our L1 bandwidth calculation.
+
+Add a link to URL explaining how to decode the hdrh data.
+
+Do we want to require more structure for dut_version, e.g. at least two dots?
+
+Should we parse version strings?
+E.g.: Turn dut_version from "21.11.0"
+into {"major": 21, "minor": 11, "patch": 0}.
diff --git a/docs/model/current/schema/yaml2json.py b/docs/model/current/schema/yaml2json.py
new file mode 100644
index 0000000000..6927bc0172
--- /dev/null
+++ b/docs/model/current/schema/yaml2json.py
@@ -0,0 +1,29 @@
+# Copyright (c) 2021 Cisco and/or its affiliates.
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at:
+#
+# http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+"""Utility to convert from .schema.yaml to .schema.json.
+
+TODO: Read the input file name from command line argument.
+"""
+
+import glob
+import json
+import yaml
+
+
+for filename in glob.glob(u"*.schema.yaml"):
+ name = filename[:-5]
+ with open(f"{name}.yaml", u"rt") as fin:
+ data = yaml.load(fin.read())
+ with open(f"{name}.json", u"wt") as fout:
+ json.dump(data, fout, indent=2)
diff --git a/docs/model/current/top.rst b/docs/model/current/top.rst
index 640abe2343..d86e3fde4c 100644
--- a/docs/model/current/top.rst
+++ b/docs/model/current/top.rst
@@ -22,7 +22,7 @@ especially the export side (UTI), not import side (PAL).
Version
~~~~~~~
-This document is valid for CSIT model version 0.1.0.
+This document is valid for CSIT model version 1.0.0.
It is recommended to use semantic versioning: https://semver.org/
That means, if the new model misses a field present in the old model,
@@ -42,7 +42,42 @@ UTI stands for Unified Test Interface.
It mainly focuses on exporting information gathered during test run
into JSON output files.
-Files
------
+Output Structure
+-----------------
-No files are exported yet in this version.
+UTI outputs come in filesystem tree structure (single tree), where directories
+correspond to suite levels and files correspond to suite setup, suite teardown
+or any test case at this level of suite.
+The directory name comes from SUITE_NAME Robot variable (the last part
+as the previous parts are higher level suites), converted to lowercase.
+If the suite name contains spaces (Robot converts underscores to spaces),
+they are replaced with underscores.
+
+The filesystem tree is rooted under tests/ (as suites in git are there),
+and for each component (test case, suite setup, suite teardown)
+two files are generated.
+The "raw" variant is suitable for debugging (can contain lower level logging),
+while the "info" variant is suitable for processing by PAL
+(can contain derivative values so PAL does not need to compute them
+on every download).
+Their structure and content is mostly identical, model definition mentions
+if a particular subschema is not identical in the two variants.
+It is possible to convert from raw to info, but not the other way.
+
+Although we expect only ASCII text in the exported files,
+we manipulate files using UTF-8 encoding,
+so if Robot Framework uses a non-ascii character, it will be handled.
+
+JSON schemas
+------------
+
+CSIT model is formally defined as a collection of JSON schema documents,
+one for each output file type.
+
+The current version specifies only one output file type:
+Info output for test case.
+
+The authoritative JSON schema documents are in JSON format.
+Git repository also contains YAML formatted document and conversion utility,
+which simplifies maintaining of the JSON document
+(no need to track brackets and commas), but are not authoritative.