{"id":{"repo_id":"vilnius","oai_identifier":"oai:vu.lt:elaba:308336427"},"canonical_url":"https://search.ndltd.org/etd/vilnius/oai:vu.lt:elaba:308336427","repository":{"repo_id":"vilnius","name":"Vilnius University","base_url":"https://epublications.vu.lt/oai"},"display":{"title":"Anglies dvideginio pėdsako optimizavimas Kubernetes valdomose sistemose /","abstract":"This bachelor's thesis investigates carbon footprint optimization possibilities in containerized systems managed by the Kubernetes orchestration platform. The aim of the work is to evaluate the suitability of different time series models for short-term electricity grid carbon intensity forecasting across different geographical zones and to assess how such forecasts could be used to reduce the environmental impact of containerized systems. Hourly electricity grid carbon intensity datasets were prepared for the United Kingdom, France, and Lithuania. The data were collected from different publicly available sources and standardized by timestamps, measurement units, and feature structure. Three forecasting models were developed and compared: XGBoost, LSTM, and VAR. The models were evaluated using MAE, RMSE, and MAPE metrics, applying a chronological split into training, validation, and testing datasets. The experimental results showed that no single model performed best across all analyzed geographical zones. In the Lithuanian dataset, XGBoost and VAR achieved very similar results, although XGBoost produced slightly more stable errors. In the French and United Kingdom datasets, the VAR model achieved the best results. The LSTM model did not achieve the lowest error in any geographical zone, although its results in the United Kingdom case were close to the best-performing model - VAR. The thesis also presents a prototype forecasting service that stores generated forecasts in a database, exposes them through an HTTP interface, and exports Prometheus metrics. A demonstrational Kubernetes scenario shows that forecasted carbon intensity values can be used as an external signal for KEDA and HPA mechanisms that adjust the number of replicas of regional workloads.","abstract_html":"This bachelor&#x27;s thesis investigates carbon footprint optimization possibilities in containerized systems managed by the Kubernetes orchestration platform. The aim of the work is to evaluate the suitability of different time series models for short-term electricity grid carbon intensity forecasting across different geographical zones and to assess how such forecasts could be used to reduce the environmental impact of containerized systems. Hourly electricity grid carbon intensity datasets were prepared for the United Kingdom, France, and Lithuania. The data were collected from different publicly available sources and standardized by timestamps, measurement units, and feature structure. Three forecasting models were developed and compared: XGBoost, LSTM, and VAR. The models were evaluated using MAE, RMSE, and MAPE metrics, applying a chronological split into training, validation, and testing datasets. The experimental results showed that no single model performed best across all analyzed geographical zones. In the Lithuanian dataset, XGBoost and VAR achieved very similar results, although XGBoost produced slightly more stable errors. In the French and United Kingdom datasets, the VAR model achieved the best results. The LSTM model did not achieve the lowest error in any geographical zone, although its results in the United Kingdom case were close to the best-performing model - VAR. The thesis also presents a prototype forecasting service that stores generated forecasts in a database, exposes them through an HTTP interface, and exports Prometheus metrics. A demonstrational Kubernetes scenario shows that forecasted carbon intensity values can be used as an external signal for KEDA and HPA mechanisms that adjust the number of replicas of regional workloads.","abstract_has_math":false,"creators":["Buzys, Justas,"],"institution":"Institutional Repository of Vilnius University","degree_name":null,"degree_level":null,"degree_discipline":null,"degree_department":null,"school":null,"contributors":[],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2026,"date_issued":"2026","date_published":"2026","updated_at":"2026-11-19T10:12:59Z","subjects":[],"languages":["lit"],"rights":["info:eu-repo/semantics/openAccess"],"rights_urls":[],"identifier_entries":[],"person_identifier_entries":[]},"links":{"outbound_url":"https://repository.vu.lt/VU:ELABAETD308336427&prefLang=en_US","outbound_label":"Repository record","outbound_source":"dc:identifier","record_url":"https://repository.vu.lt/VU:ELABAETD308336427&prefLang=en_US","record_label":"Repository record","record_kind":"record","record_source":"dc:identifier","download_url":null,"download_label":null,"download_source":null,"access_status":"open","embargo_until":null,"access_note":null},"source_record":{"url":"https://epublications.vu.lt/oai?verb=GetRecord&metadataPrefix=oai_dc&identifier=oai%3Avu.lt%3Aelaba%3A308336427","prefix":"oai_dc"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:creator","label":"Author","values":["Buzys, Justas,"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2026"]},{"key":"dc:publisher","label":"Institution","values":["Institutional Repository of Vilnius University"]},{"key":"dc:relation","label":"Dc Relation","values":["https://epublications.vu.lt/object/elaba:308336427/308336427.pdf"]},{"key":"dc:type","label":"Dc Type","values":["info:eu-repo/semantics/bachelorThesis"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["lit"]},{"key":"dc:rights","label":"Dc Rights","values":["info:eu-repo/semantics/openAccess"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["https://repository.vu.lt/VU:ELABAETD308336427&prefLang=en_US"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["This bachelor's thesis investigates carbon footprint optimization possibilities in containerized systems managed by the Kubernetes orchestration platform. The aim of the work is to evaluate the suitability of different time series models for short-term electricity grid carbon intensity forecasting across different geographical zones and to assess how such forecasts could be used to reduce the environmental impact of containerized systems. Hourly electricity grid carbon intensity datasets were prepared for the United Kingdom, France, and Lithuania. The data were collected from different publicly available sources and standardized by timestamps, measurement units, and feature structure. Three forecasting models were developed and compared: XGBoost, LSTM, and VAR. The models were evaluated using MAE, RMSE, and MAPE metrics, applying a chronological split into training, validation, and testing datasets. The experimental results showed that no single model performed best across all analyzed geographical zones. In the Lithuanian dataset, XGBoost and VAR achieved very similar results, although XGBoost produced slightly more stable errors. In the French and United Kingdom datasets, the VAR model achieved the best results. The LSTM model did not achieve the lowest error in any geographical zone, although its results in the United Kingdom case were close to the best-performing model - VAR. The thesis also presents a prototype forecasting service that stores generated forecasts in a database, exposes them through an HTTP interface, and exports Prometheus metrics. A demonstrational Kubernetes scenario shows that forecasted carbon intensity values can be used as an external signal for KEDA and HPA mechanisms that adjust the number of replicas of regional workloads."]},{"key":"dc:format","label":"Dc Format","values":["application/pdf"]},{"key":"dc:title","label":"Title","values":["Anglies dvideginio pėdsako optimizavimas Kubernetes valdomose sistemose /","Carbon footprint optimization in kubernetes-managed systems."]}]}],"canonical_facts":{"dc:creator":["Buzys, Justas,"],"dc:date":["2026"],"dc:description":["This bachelor's thesis investigates carbon footprint optimization possibilities in containerized systems managed by the Kubernetes orchestration platform. The aim of the work is to evaluate the suitability of different time series models for short-term electricity grid carbon intensity forecasting across different geographical zones and to assess how such forecasts could be used to reduce the environmental impact of containerized systems. Hourly electricity grid carbon intensity datasets were prepared for the United Kingdom, France, and Lithuania. The data were collected from different publicly available sources and standardized by timestamps, measurement units, and feature structure. Three forecasting models were developed and compared: XGBoost, LSTM, and VAR. The models were evaluated using MAE, RMSE, and MAPE metrics, applying a chronological split into training, validation, and testing datasets. The experimental results showed that no single model performed best across all analyzed geographical zones. In the Lithuanian dataset, XGBoost and VAR achieved very similar results, although XGBoost produced slightly more stable errors. In the French and United Kingdom datasets, the VAR model achieved the best results. The LSTM model did not achieve the lowest error in any geographical zone, although its results in the United Kingdom case were close to the best-performing model - VAR. The thesis also presents a prototype forecasting service that stores generated forecasts in a database, exposes them through an HTTP interface, and exports Prometheus metrics. A demonstrational Kubernetes scenario shows that forecasted carbon intensity values can be used as an external signal for KEDA and HPA mechanisms that adjust the number of replicas of regional workloads."],"dc:format":["application/pdf"],"dc:identifier":["https://repository.vu.lt/VU:ELABAETD308336427&prefLang=en_US"],"dc:language":["lit"],"dc:publisher":["Institutional Repository of Vilnius University"],"dc:relation":["https://epublications.vu.lt/object/elaba:308336427/308336427.pdf"],"dc:rights":["info:eu-repo/semantics/openAccess"],"dc:title":["Anglies dvideginio pėdsako optimizavimas Kubernetes valdomose sistemose /","Carbon footprint optimization in kubernetes-managed systems."],"dc:type":["info:eu-repo/semantics/bachelorThesis"]},"updated_at":"2026-11-19T10:12:59Z"}