From owner-freebsd-isp Tue Feb 29 8:53:15 2000 Delivered-To: freebsd-isp@freebsd.org Received: from flashmail.com (flash1.flashmail.com [207.173.216.240]) by hub.freebsd.org (Postfix) with SMTP id 4A67437B510 for ; Tue, 29 Feb 2000 08:53:12 -0800 (PST) (envelope-from mholloway@flashmail.com) Received: from sna ([206.135.117.1]) by flashmail.com ; Tue, 29 Feb 2000 08:53:19 -0800 Message-ID: <002001bf82d5$7f5a6a20$a52410ac@sierrahealth.com> From: "Mark Holloway" To: Subject: OC3 versus T1 Circuits Date: Tue, 29 Feb 2000 08:53:25 -0800 MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_NextPart_000_001D_01BF8292.709E6C90" X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.00.2314.1300 X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2314.1300 Sender: owner-freebsd-isp@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org This is a multi-part message in MIME format. ------=_NextPart_000_001D_01BF8292.709E6C90 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable I have a situation and maybe some of you can please advise: I have a core LAN/MAN/WAN campus with approximately 80 servers. I have = about ten different remote sites throughout the city (the MAN) which = clients log into a Windows NT domain and then access certain = applications. Until late 1999 they were running these applications in a = client/server fashion. The ten sites are all on a shared FDDI ring, but = each location is a 10MB, shared, half duplex connection. The original = strategy was to have a full OC3 from the main campus going to a Sprint = Central Office, then have 10MB fractional OC3 going to each site (almost = like Frame Relay in the MAN). However, we have since setup many Windows = Terminal Servers (25 servers @ 200 clients per server) and the clients = are using Citrix on their local desktops. This solution works well. = But now I am wondering if the fractional OC3 is overkill?? I was = thinking maybe either a T1 line or two T1 lines bonded for EACH SITE = rather than a 10MB OC3 for each site would be more realistic? Is a T1 = really .15 MB? Or 1.5MB? I think the slowness that most people = experience is due to the nature of the FDDI. Each site averages about = 60 clients, but a couple have up to 150 clients. When using Citrix = everything runs fine. The only apps they would run locally are Outlook = and some telnet sessions (pure ANSI, little overhead). =20 I apologize if this is too off topic, but I've always tried to = contribute to this list whenever possible. One thing to keep in mind is = that for each OC3 remote connection we were going to buy a 3Com = Pathbuilder 330 (designed for fractional OC3). This is approximately = $12,000 + the Pathbuilder 700 Ethernet blade for the the WAN switch at = the main campus (another several thousand dollars). A Cisco 2500 or = 2600 with bonded T1 is under $2000. =20 PLEASE, if anyone has any insite, feedback, or comments, I'd really = appreciate it.=20 Regards, Mark ------=_NextPart_000_001D_01BF8292.709E6C90 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable

I have a situation and maybe some of = you can please=20 advise:

I have a core LAN/MAN/WAN campus with = approximately=20 80 servers. I have about ten different remote sites throughout the = city=20 (the MAN) which clients log into a Windows NT domain and then access = certain=20 applications. Until late 1999 they were running these applications = in a=20 client/server fashion. The ten sites are all on a shared FDDI = ring, but=20 each location is a 10MB, shared, half duplex connection. The = original=20 strategy was to have a full OC3 from the main campus going to a Sprint = Central=20 Office, then have 10MB fractional OC3 going to each site (almost like = Frame=20 Relay in the MAN). However, we have since setup many Windows = Terminal=20 Servers (25 servers @ 200 clients per server) and the clients are using = Citrix=20 on their local desktops. This solution works well. But now I = am=20 wondering if the fractional OC3 is overkill?? I was thinking maybe = either=20 a T1 line or two T1 lines bonded for EACH SITE rather than a 10MB OC3 = for each=20 site would be more realistic? Is a T1 really .15 MB? Or=20 1.5MB? I think the slowness that most people experience is = due to=20 the nature of the FDDI. Each site averages about 60 clients, but a = couple=20 have up to 150 clients. When using Citrix everything runs = fine. The=20 only apps they would run locally are Outlook and some telnet sessions = (pure=20 ANSI, little overhead).

I apologize if this is too off topic, = but I've=20 always tried to contribute to this list whenever possible. One = thing to=20 keep in mind is that for each OC3 remote connection we were going to buy = a 3Com=20 Pathbuilder 330 (designed for fractional OC3). This is = approximately=20 $12,000 + the Pathbuilder 700 Ethernet blade for the the WAN switch at = the main=20 campus (another several thousand dollars). A Cisco 2500 or 2600 = with=20 bonded T1 is under $2000.

PLEASE, if anyone has any insite, = feedback, or=20 comments, I'd really appreciate it.

Regards,

Mark

------=_NextPart_000_001D_01BF8292.709E6C90-- To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-isp" in the body of the message