- Community Home
- >
- Servers and Operating Systems
- >
- ProLiant
- >
- ProLiant Servers (ML,DL,SL)
- >
- Compaq NIC "Speed & Duplex" auto configuration
-
- Forums
-
Blogs
- Alliances
- Around the Storage Block
- Behind the scenes @ Labs
- HPE Careers
- HPE Storage Tech Insiders
- Infrastructure Insights
- Inspiring Progress
- Internet of Things (IoT)
- My Learning Certification
- OEM Solutions
- Servers: The Right Compute
- Shifting to Software-Defined
- Telecom IQ
- Transforming IT
- Infrastructure Solutions German
- L’Avenir de l’IT
- IT e Trasformazione Digitale
- Enterprise Topics
- ИТ для нового стиля бизнеса
- Blogs
-
Quick Links
- Community
- Getting Started
- FAQ
- Ranking Overview
- Rules of Participation
- Contact
- Email us
- Tell us what you think
- Information Libraries
- Integrated Systems
- Networking
- Servers
- Storage
- Other HPE Sites
- Support Center
- Enterprise.nxt
- Marketplace
- Aruba Airheads Community
-
Forums
-
Blogs
-
InformationEnglish
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page
- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content
09-30-2003 05:21 AM
09-30-2003 05:21 AM
We experienced connection stability problem when connecting to an Oracle database on an Oracle Server (Windows 2000 server, Compaq ML370 with NC7770 on Cisco catalyst 4006 gigabit switch. Some of the client are connected through a WAN linked to the 4006).
The connection to the server does work OK but is interrupted from time to time for a few seconds. These interruptions do not happen regularly, but more often when the network is busy. They can happen once a day or twice in half an hour and have a duration of a few seconds up to 1 minute. As every client are disconnected (Oracle connection stalled) at the same time, the problem obviously is on the server side.
Note that removing the network cable for 2 minutes on the client side is supported without network interruptions (the TCP/IP network configuration has been set accordingly).
Can it be a problem with the current "Speed & Duplex" configuration that is set to Auto?
This diagnostic is based on the many problems that are related to auto speed and duplex configuration that can be found in users forums, and on the fact that by manually changing the duplex mode or the interface speed the connections are immediately broken (it is not needed to wait on the TCP timeout connection as when unplugging the network cable).
Questions are:
- How does the adapter decide to change its duplex mode or speed settings?
- Can the adapter behavior be configured? i.e. is it possible to increase the number of connection retry before the adapter decide to auto-reconfigurate?
- Is it possible to force a Gigabit speed setting for NC7xxx series cards? Only 10 and 100 Mb/s connection speed are listed.
Thanks in advance for your ideas, experiences, inputs, or even better your answers.
Patrick
Solved! Go to Solution.
- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content
09-30-2003 12:39 PM
09-30-2003 12:39 PM
Re: Compaq NIC "Speed & Duplex" auto configuration
We've certainly experienced problems with Auto Negoitation and have fixed all our Servers at both the Server and Network end.
We found cards would set to inappropriate speeds and half duplex which really had an impact on performance.
I don't know what can trigger an re-autonegotiation process.
- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content
10-02-2003 01:11 AM
10-02-2003 01:11 AM
Re: Compaq NIC "Speed & Duplex" auto configuration
- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content
10-02-2003 02:09 AM
10-02-2003 02:09 AM
SolutionWe experienced a very similar problem with a Proliant ML580, installed with SmartStart 5.40 ( and running during months perfectly ).
In our case, that strange behaviour was due to NIC's & teaming drivers version. We updated to 6.20 version and automagically all was OK.
And other Proliant same model, same Smartart, same installation date, same function, was running OK all the time.
Kind regards,
Zigor
Hewlett Packard Enterprise International
- Communities
- HPE Blogs and Forum
© Copyright 2019 Hewlett Packard Enterprise Development LP