Aus IPD-Institutsseminar
Version vom 19. Juni 2018, 12:16 Uhr von Erik Burger (Diskussion | Beiträge) (Die Seite wurde neu angelegt: „{{Termin |datum=2018/12/21 11:30:00 |raum=Raum 348 (Gebäude 50.34) }}“)
(Unterschied) ← Nächstältere Version | Aktuelle Version (Unterschied) | Nächstjüngere Version → (Unterschied)
Zur Navigation springen Zur Suche springen
Termin (Alle Termine)
Datum Fr 21. Dezember 2018, 11:30 Uhr
Dauer 65 min
Ort Raum 348 (Gebäude 50.34)
Vorheriger Termin Fr 14. Dezember 2018
Nächster Termin Fr 11. Januar 2019


Vortragende(r) Erdinch Mehmedali
Titel Metric-driven job assignment in cloud-based environments
Vortragstyp Masterarbeit
Betreuer(in) Robert Heinrich
Kurzfassung A cloud storage migration is usually done in one of two ways - via corresponding storage sync clients or SaaS migration tools. The SaaS migration tools can typically migrate the data significantly faster, as they are not as constrained by the Internet bandwidth as the users are. Such tools incorporate a server that reads the data from the user’s old cloud storage and copies it to another, desired cloud storage. This server is referred to as "migration server". The geographic location of the migration server can influence the duration of the cloud storage migration. Commonly, it is reported that closer distances yield better results in terms of speed over the Internet and hence, the expectation is that a migration server placed at the geographic midpoint between the data centers of the cloud storages involved, will lead to good results. We investigate different influences on cloud storage migration speed and conceptualize an algorithm for choosing a migration server location in a generic cloud storage migration scenario. In an experimental evaluation, the results of the algorithm are compared against the results of the geographic midpoint between the data centers of the cloud storages involved in the migration. midpoint, determine the necessity of developing an algorithm for choosing a migration serverlocation and ultimately present some of the guidelines for developing such an algorithm.
Vortragende(r) Nicolas Boltz
Titel State of the Art: Multi Actor Behaviour and Dataflow Modelling for Dynamic Privacy
Vortragstyp Proposal
Betreuer(in) Robert Heinrich
Kurzfassung State of the Art Vortrag im Rahmen der Praxis der Forschung.
Neuen Vortrag erstellen