cypherean commented on code in PR #14250: URL: https://github.com/apache/pinot/pull/14250#discussion_r2146926125
########## pinot-controller/src/main/java/org/apache/pinot/controller/api/resources/PinotSegmentRestletResource.java: ########## @@ -926,7 +926,8 @@ public String getServerMetadata( @ApiParam(value = "Name of the table", required = true) @PathParam("tableName") String tableName, @ApiParam(value = "OFFLINE|REALTIME") @QueryParam("type") String tableTypeStr, @ApiParam(value = "Columns name", allowMultiple = true) @QueryParam("columns") @DefaultValue("") - List<String> columns, @Context HttpHeaders headers) { + List<String> columns, @ApiParam(value = "Segments name", allowMultiple = true) @QueryParam("segments") Review Comment: This is in sync with the column arg, where the params are decoded only once in the servers. Do you think we need to decode it here? ########## pinot-controller/src/main/java/org/apache/pinot/controller/util/TableMetadataReader.java: ########## @@ -127,50 +130,111 @@ private TableReloadJsonResponse processSegmentMetadataReloadResponse( /** * This api takes in list of segments for which we need the metadata. + * This calls the server to get the metadata for all segments instead of making a call per segment. */ public JsonNode getSegmentsMetadata(String tableNameWithType, List<String> columns, Set<String> segmentsToInclude, int timeoutMs) throws InvalidConfigException, IOException { return getSegmentsMetadataInternal(tableNameWithType, columns, segmentsToInclude, timeoutMs); } - private JsonNode getSegmentsMetadataInternal(String tableNameWithType, List<String> columns, - Set<String> segmentsToInclude, int timeoutMs) + /** + * Common helper used by both the new (table-level) and legacy + * (segment-level) endpoints. + */ + private JsonNode fetchAndAggregateMetadata(List<String> urls, + BiMap<String, String> endpoints, + boolean perSegmentJson, + String tableNameWithType, + int timeoutMs) throws InvalidConfigException, IOException { - final Map<String, List<String>> serverToSegmentsMap = - _pinotHelixResourceManager.getServerToSegmentsMap(tableNameWithType); - BiMap<String, String> endpoints = - _pinotHelixResourceManager.getDataInstanceAdminEndpoints(serverToSegmentsMap.keySet()); - ServerSegmentMetadataReader serverSegmentMetadataReader = - new ServerSegmentMetadataReader(_executor, _connectionManager); - // Filter segments that we need - for (Map.Entry<String, List<String>> serverToSegment : serverToSegmentsMap.entrySet()) { - List<String> segments = serverToSegment.getValue(); - if (segmentsToInclude != null && !segmentsToInclude.isEmpty()) { - segments.retainAll(segmentsToInclude); + CompletionServiceHelper cs = + new CompletionServiceHelper(_executor, _connectionManager, endpoints); + CompletionServiceHelper.CompletionServiceResponse resp = + cs.doMultiGetRequest(urls, tableNameWithType, perSegmentJson, timeoutMs); + + if (resp._failedResponseCount == urls.size()) { + throw new InvalidConfigException("All requests to server instances failed."); + } + + ObjectMapper mapper = new ObjectMapper(); + ObjectNode aggregatedNode = mapper.createObjectNode(); + + for (String body : resp._httpResponses.values()) { + JsonNode node = JsonUtils.stringToJsonNode(body); + + // legacy returns one JSON per segment; new returns one JSON with many fields + if (perSegmentJson) { + String segmentName = node.get("segmentName").asText(); + aggregatedNode.set(segmentName, node); + } else { + node.fields().forEachRemaining(entry -> aggregatedNode.set(entry.getKey(), entry.getValue())); } } + return aggregatedNode; + } - List<String> segmentsMetadata = - serverSegmentMetadataReader.getSegmentMetadataFromServer(tableNameWithType, serverToSegmentsMap, endpoints, - columns, timeoutMs); - Map<String, JsonNode> response = new HashMap<>(); - for (String segmentMetadata : segmentsMetadata) { - JsonNode responseJson = JsonUtils.stringToJsonNode(segmentMetadata); - response.put(responseJson.get("segmentName").asText(), responseJson); + private List<String> buildTableLevelUrls(Map<String, List<String>> serverToSegs, + BiMap<String, String> endpoints, + String tableNameWithType, + List<String> columns, + Set<String> segmentsFilter, + ServerSegmentMetadataReader reader) { + + List<String> urls = new ArrayList<>(serverToSegs.size()); + for (String server : serverToSegs.keySet()) { + urls.add(reader.generateTableMetadataServerURL( + tableNameWithType, columns, segmentsFilter, endpoints.get(server))); } - return JsonUtils.objectToJsonNode(response); + return urls; } - /** - * This method retrieves the full segment metadata for a given table. - * Currently supports only OFFLINE tables. - * @return a map of segmentName to its metadata - */ - public JsonNode getSegmentsMetadata(String tableNameWithType, List<String> columns, int timeoutMs) + private List<String> buildSegmentLevelUrls(Map<String, List<String>> serverToSegs, + BiMap<String, String> endpoints, + String tableNameWithType, + List<String> columns, + Set<String> segmentsFilter, + ServerSegmentMetadataReader reader) { + + List<String> urls = new ArrayList<>(); + for (Map.Entry<String, List<String>> e : serverToSegs.entrySet()) { + for (String segment : e.getValue()) { + if (segmentsFilter == null || segmentsFilter.isEmpty() + || segmentsFilter.contains(segment)) { + urls.add(reader.generateSegmentMetadataServerURL( + tableNameWithType, segment, columns, endpoints.get(e.getKey()))); + } + } + } + return urls; + } + + private JsonNode getSegmentsMetadataInternal(String tableNameWithType, List<String> columns, + Set<String> segmentsToInclude, int timeoutMs) throws InvalidConfigException, IOException { - return getSegmentsMetadataInternal(tableNameWithType, columns, null, timeoutMs); + Map<String, List<String>> serverToSegs = + _pinotHelixResourceManager.getServerToSegmentsMap(tableNameWithType); + BiMap<String, String> endpoints = + _pinotHelixResourceManager.getDataInstanceAdminEndpoints(serverToSegs.keySet()); + ServerSegmentMetadataReader reader = + new ServerSegmentMetadataReader(_executor, _connectionManager); + + // try table level endpoint first + try { + List<String> tableUrls = buildTableLevelUrls(serverToSegs, endpoints, + tableNameWithType, columns, segmentsToInclude, reader); + return fetchAndAggregateMetadata(tableUrls, endpoints, /*perSegmentJson=*/false, + tableNameWithType, timeoutMs); + } catch (InvalidConfigException ignore) { Review Comment: `fetchAndAggregateMetadata()` throw `InvalidConfigException` when all server requests fail which is expected when new server API doesn't exist. We ignore and fallback to the legacy API in this case. The function also throws `IOException` when stringToJsonNode conversion fails. This exception is not caught and propagated as is. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: commits-unsubscr...@pinot.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: commits-unsubscr...@pinot.apache.org For additional commands, e-mail: commits-h...@pinot.apache.org