Good morning to all, I am coordinating a project for installing a wireless network at the Faculty of Statistics of University of Rome. We have been using freeradius with PEAP and EAP-TTLS enabled. All seems to work fine except for the very long authentication times (about one minutes) we are experiencing when using the default Windows XP supplicant. After doing some searches on the previous discussions, I found the following thread where it is said that such a problem may be due to the way the Windows supplicant handles the EAP negotiation: http://www.mail-archive.com/freeradius-users@lists.freeradius.org/msg30473.h... I performed some further searches and discovered that this delay may be due to the four-way handshake protocol used by the client machine to establish a new connection with the AP. Do someone knows if there exists any tuning in the implementation of this protocol able to reduce the waiting time? Thanks in advance for any help! -- Umberto Ferraro Petrillo PhD Dipartimento di Statistica, Probabilita' e Statistiche Applicate Universita' di Roma -- La Sapienza P.le Aldo Moro, 5 00185 - Rome, Italy phone: +39 06 499 10513 e-mail: Umberto.Ferraro@uniroma1.it
Good evening,
Good morning to all, I am coordinating a project for installing a wireless network at the Faculty of Statistics of University of Rome. We have been using freeradius with PEAP and EAP-TTLS enabled. All seems to work fine except for the very long authentication times (about one minutes) we are experiencing when using the default Windows XP supplicant.
After doing some searches on the previous discussions, I found the following thread where it is said that such a problem may be due to the way the Windows supplicant handles the EAP negotiation:
http://www.mail-archive.com/freeradius-users@lists.freeradius.org/msg30473. html
I performed some further searches and discovered that this delay may be due to the four-way handshake protocol used by the client machine to establish a new connection with the AP. Do someone knows if there exists any tuning in the implementation of this protocol able to reduce the waiting time?
Thanks in advance for any help!
I haven't done any extensive research, but my bundled Intel PRO/Wireless supplicant (comes with Centrino chipsets) gets it done within a few seconds. I believe that the time of issuing the DHCP request after authentication is quite crucial. If it is triggered too early, you hang around in an authenticated-but-not-ready state until the DHCP request times out and is tried again. Greetings, Stefan -- Stefan WINTER Stiftung RESTENA - Réseau Téléinformatique de l'Education Nationale et de la Recherche Ingenieur Forschung & Entwicklung 6, rue Richard Coudenhove-Kalergi L-1359 Luxembourg E-Mail: stefan.winter@restena.lu Tel.: +352 424409-1 http://www.restena.lu Fax: +352 422473
Good morning Stefan, perhaps the solution you proposed could work in my case. Could you suggest me the way to check if the lag depends on the DHCP service? Do you know if there is any way to configure the DHCP service timing (we are using freeradius + Chillispot on the server side) so to avoid this case? Thank you! On 12/18/06, Stefan Winter <stefan.winter@restena.lu> wrote:
Good evening,
Good morning to all, I am coordinating a project for installing a wireless network at the Faculty of Statistics of University of Rome. We have been using freeradius with PEAP and EAP-TTLS enabled. All seems to work fine except for the very long authentication times (about one minutes) we are experiencing when using the default Windows XP supplicant. ...
Thanks in advance for any help!
I haven't done any extensive research, but my bundled Intel PRO/Wireless supplicant (comes with Centrino chipsets) gets it done within a few seconds. I believe that the time of issuing the DHCP request after authentication is quite crucial. If it is triggered too early, you hang around in an authenticated-but-not-ready state until the DHCP request times out and is tried again.
Greetings,
Stefan
-- Umberto Ferraro Petrillo PhD Dipartimento di Statistica, Probabilita' e Statistiche Applicate Universita' di Roma -- La Sapienza P.le Aldo Moro, 5 00185 - Rome, Italy phone: +39 06 499 10513 e-mail: Umberto.Ferraro@uniroma1.it
Hi,
Good morning Stefan, perhaps the solution you proposed could work in my case. Could you suggest me the way to check if the lag depends on the DHCP service? Do you know if there is any way to configure the DHCP service timing (we are using freeradius + Chillispot on the server side) so to avoid this case?
Well, this is almost exclusively a client-side issue. From FreeRADIUS perspective, the job's done as soon as the authentication succeeded. The issue is that the client may try to ask for an IP address before the port is actually opened, and I don't know how to influence timing on an (XP?) client. The only option I see here is for chillispot: it might act a bit more gratious when receiving a DHCP request but the port has just not been opened - it might let this request through anyways. Or log it, and only answer to it after it has opened the port. In any case nothing where the freeradius-users mailing list could help you, sorry. Stefan -- Stefan WINTER Stiftung RESTENA - Réseau Téléinformatique de l'Education Nationale et de la Recherche Ingenieur Forschung & Entwicklung 6, rue Richard Coudenhove-Kalergi L-1359 Luxembourg E-Mail: stefan.winter@restena.lu Tel.: +352 424409-1 http://www.restena.lu Fax: +352 422473
Egr.Sig Ferraro, mi permetto di scriverle in italiano, visto che non è mia intenzione aggiungere informazioni di carattere tecnico, lasciando tale compito agli sviluppatori di Freeradius certamente più competenti, tuttavia, a proposito del progetto che state sviluppando, desidero presentarle la ditta per cui lavoro che si occupa proprio di realizzare reti wireless e gestire gli accessi in modo centralizzato. La Guglielmo (questo è il nome della ditta in onore a Marconi) copre già con i suoi servizi biblioteche comunali e universitarie, hotel, strutture tuistiche e centri urbani. Il completo controllo del sistema di autenticazione centralizzato permette di adattare il servizio dal punto di vista grafico-promozionale alla realtà dove viene attivato, mantenendo la compatibilità tra tutte le location, in questo modo un utente può usare il proprio account in ognuno dei nostri punti. Il sistema centralizzato permette l'archiviazione dei deti relativi agli utenti e alle connessioni effettuete secondo le norme vigenti in modo automatico e l'inserimento di ogni nuova location in un circuito sempre più ampio, arricchito da accordi di roaming con altre compagnie straniere. Può avere altre informazioni consultando : www.guglielmo.biz Se desidera approfondire la conoscenza reciproca non esiti a contattarmi al : nicola.iotti@guglielmo.biz Mi perdoni se mi sono permesso di scrivere senza conoscela e senza conoscere i dettagli del progetto a cui sta lavorando, ma la mia unica intenzione è quella di provare a mettere in contatto realtà per verificare eventuali possibilità di collaborazione. Cordiali Saluti -------Original Message------- From: Umberto.ferraro@uniroma1.it; FreeRadius users mailing list Date: 12/18/06 16:44:11 To: freeradius-users@lists.freeradius.org Subject: Long authentication times Good morning to all, I am coordinating a project for installing a wireless network at the Faculty of Statistics of University of Rome. We have been using freeradius with PEAP and EAP-TTLS enabled. All seems to work fine except for the very long authentication times (about one minutes) we are experiencing when using the default Windows XP supplicant. After doing some searches on the previous discussions, I found the following thread where it is said that such a problem may be due to the way the Windows supplicant handles the EAP negotiation: http://www.mail-archive.com/freeradius-users@lists.freeradius.org/msg30473 html I performed some further searches and discovered that this delay may be due to the four-way handshake protocol used by the client machine to establish a new connection with the AP. Do someone knows if there exists any tuning in the implementation of this protocol able to reduce the waiting time? Thanks in advance for any help! -- Umberto Ferraro Petrillo PhD Dipartimento di Statistica, Probabilita' e Statistiche Applicate Universita' di Roma -- La Sapienza P.le Aldo Moro, 5 00185 - Rome, Italy phone: +39 06 499 10513 e-mail: Umberto.Ferraro@uniroma1.it - List info/subscribe/unsubscribe? See http://www.freeradius.org/list/users html
participants (3)
-
Nicola Iotti -
Stefan Winter -
Umberto Ferraro Petrillo