Hi all,
We have a problem for months now.
Open-E doesn't help us very much and the hardware vendor says it's normal...
I hope someone can help us. (also a fee is not a problem)
What's the problem?
We running every hour a 'BDCC CHECKDB' at MSSQL 2008 R2 (newest version).
Sometimes we get inconsistency errors.
We test it with a DB at the Open-E datastore and with a DB at a USB disk.
The USB disk runs everytime good, at the datastore not.
When we running with only the SQL VM it runs fine
when we running with more VM's we get inconsistency error's.
Sometimes with 2 or 3 VM's it also runs good.
What a kind of envoirment we running?
-VMWare 5.10 build 1900470
-Open-E v6
Version: 6.0up99.8101.11764 64bit
Release date: 2014-05-16
The 5 VMWare hosts are directly connected to the datastore with 10GBPs Intel networkadapters.
Here all the adapters in the datastore:
Intel Corporation 82576 Gigabit Network Connection (rev 01)
Intel Corporation 82576 Gigabit Network Connection (rev 01)
Intel Corporation 82599ES 10-Gigabit SFI/SFP+ Network Connection (rev 01)
Intel Corporation 82599ES 10-Gigabit SFI/SFP+ Network Connection (rev 01)
Intel Corporation 82599ES 10-Gigabit SFI/SFP+ Network Connection (rev 01)
Intel Corporation 82599ES 10-Gigabit SFI/SFP+ Network Connection (rev 01)
Intel Corporation 82599ES 10-Gigabit SFI/SFP+ Network Connection (rev 01)
Intel Corporation 82599ES 10-Gigabit SFI/SFP+ Network Connection (rev 01)
We working with two of the same datastores with failover function.
No strange things, in the logs only some warning messages with the following warning Failover: Ping node 192.168.x.x is down
But Open-E and hardware vendor says that this warnings are some not to worry about.
What have we already done?
- change 10GBPs path to 1GBPs path, exactly the same errors
- upgrade Open-E datastores to the newest version (that we get from Open-E)
- upgrade VMWare 5.1 to the latest build
- upgrade the networkcard drivers at the host themselve
- disable TCP Offload at the datastores
http://kb.open-e.com/How-to-get-better-performance-with-disabling-the-offload-option-in-NIC_1753.htm...
After this it looks like everthing runs better, it was know possible to run 3 vm's at one host without any fault.
That's the reason that we add some more VM's, but after this again errors.
- disable TCP Offload at on of the VMWare hosts
Still the same problems.
I don't know what to do. Is there sameone that can help me, we also agree with some company that can fix this for a fee.
When their some questions, ask me.
