It seems that according to
https://access.redhat.com/documentation/en-us/red_hat_virtualization/4.0/html/installation_guide/sect-firewalls
the ports of interest are:
16514/TCP
49152 - 49216/TCP
Best Regards,
Strahil NikolovOn Aug 25, 2019 08:48, Strahil wrote:
>
> Curtis,
>
> Do you have
Curtis,
Do you have enough space to run tcpdump (port not 22) on both hosts and on the
small VM you have done previously - and then start the migration?
Best Regards,
Strahil NikolovOn Aug 24, 2019 22:15, "Curtis E. Combs Jr."
wrote:
>
> I applied a 90Mbs QOS Rate Limit with 10 set for the
Curtis,
Do you have enough space to run tcpdump (port not 22) on both hosts and on the
small VM you have done previously - and then start the migration?
Best Regards,
Strahil NikolovOn Aug 24, 2019 22:15, "Curtis E. Combs Jr."
wrote:
>
> I applied a 90Mbs QOS Rate Limit with 10 set for the
Well, this proves that the issue is not the bandwidth usage, but something
else.
My personal opinion is that you should change the colocation - if that is an
option at all ...
Best Regards,
Strahil Nikolov В събота, 24 август 2019 г., 22:16:18 ч. Гринуич+3, Curtis
E. Combs Jr. написа:
I applied a 90Mbs QOS Rate Limit with 10 set for the shares to both
interfaces of 2 of the hosts. My hosts names are swm-01 and swm-02.
Creating a small VM from a Cinder template and running it gave me a test VM.
When I migrated from swm-01 to swm-02, swm-01 immediately became
unresponsive to
What is your bandwidth threshold for the network used for VM migration ?
Can you set a 90 mbit/s threshold (yes, less than 100mbit/s) and try to migrate
a small (1 GB RAM) VM ?
Do you see disconnects ?
If no, try a little bit up (the threshold) and check again.
Best Regards,
Strahil NikolovOn
On Fri, Aug 23, 2019 at 8:50 PM Tony Pearce wrote:
>
> Is the nic to the network staying up or going down for a period?
Which nic? The one on the pserver or the virtual machine? For clarity,
I've only ever referred to the one on the pserver. I can't even reach
the VM when the pserver becomes
Is the nic to the network staying up or going down for a period?
I'm just thinking, if the network has been configured to block unknown
unicast traffic, I think the VM would need to send a layer 2 frame to the
network before the network would send any frames to that switch port
destined for the
It took a while for my servers to come back on the network this time.
I think it's due to ovirt continuing to try to migrate the VMs around
like I requested. The 3 servers' names are "swm-01, swm-02 and
swm-03". Eventually (about 2-3 minutes ago) they all came back online.
So I disabled and
On Fri, Aug 23, 2019 at 9:19 PM Dominik Holler wrote:
>
>
> On Fri, Aug 23, 2019 at 8:03 PM Curtis E. Combs Jr.
> wrote:
>
>> This little cluster isn't in production or anything like that yet.
>>
>> So, I went ahead and used your ethtool commands to disable pause
>> frames on both interfaces of
On Fri, Aug 23, 2019 at 8:03 PM Curtis E. Combs Jr.
wrote:
> This little cluster isn't in production or anything like that yet.
>
> So, I went ahead and used your ethtool commands to disable pause
> frames on both interfaces of each server. I then, chose a few VMs to
> migrate around at random.
This little cluster isn't in production or anything like that yet.
So, I went ahead and used your ethtool commands to disable pause
frames on both interfaces of each server. I then, chose a few VMs to
migrate around at random.
swm-02 and swm-03 both went out again. Unreachable. Can't ping, can't
On Fri, Aug 23, 2019 at 6:45 PM Curtis E. Combs Jr.
wrote:
> Unfortunately, I can't check on the switch. Trust me, I've tried.
> These servers are in a Co-Lo and I've put 5 tickets in asking about
> the port configuration. They just get ignored - but that's par for the
> coarse for IT here. Only
Unfortunately, I can't check on the switch. Trust me, I've tried.
These servers are in a Co-Lo and I've put 5 tickets in asking about
the port configuration. They just get ignored - but that's par for the
coarse for IT here. Only about 2 out of 10 of our tickets get any
response and usually the
On Fri, Aug 23, 2019 at 5:49 PM Curtis E. Combs Jr.
wrote:
> Sure! Right now, I only have a 500gb partition on each node shared over
> NFS, added as storage domains. This is on each node - so, currently 3.
>
> How can the storage cause a node to drop out?
>
>
Thanks, I got it.
All three links go
Sure! Right now, I only have a 500gb partition on each node shared over
NFS, added as storage domains. This is on each node - so, currently 3.
How can the storage cause a node to drop out?
On Fri, Aug 23, 2019, 11:46 AM Dominik Holler wrote:
>
>
> On Fri, Aug 23, 2019 at 5:41 PM Curtis E.
On Fri, Aug 23, 2019 at 5:41 PM Curtis E. Combs Jr.
wrote:
> Also, if it helps, the hosts will sit there, quietly, for hours or
> days before anything happens. They're up and working just fine. But
> then, when I manually migrate a VM from one host to another, they
> become completely
Also, if it helps, the hosts will sit there, quietly, for hours or
days before anything happens. They're up and working just fine. But
then, when I manually migrate a VM from one host to another, they
become completely inaccessible.
These are vanilla-as-possible CentOS7 nodes. Very basic ovirt
Hey Dominik,
Thanks for helping. I really want to try to use ovirt.
When these events happen, I cannot even SSH to the nodes due to the
link being down. After a little while, the hosts come back...
On Fri, Aug 23, 2019 at 11:30 AM Dominik Holler wrote:
>
> Is you storage connected via NFS?
>
Is you storage connected via NFS?
Can you manually access the storage on the host?
On Fri, Aug 23, 2019 at 5:19 PM Curtis E. Combs Jr.
wrote:
> Sorry to dead bump this, but I'm beginning to suspect that maybe it's
> not STP that's the problem.
>
> 2 of my hosts just went down when a few VMs
Sorry to dead bump this, but I'm beginning to suspect that maybe it's
not STP that's the problem.
2 of my hosts just went down when a few VMs tried to migrate.
Do any of you have any idea what might be going on here? I don't even
know where to start. I'm going to include the dmesg in case it
Thanks, I'm just going to revert back to bridges.
On Thu, Aug 22, 2019 at 11:50 AM Dominik Holler wrote:
>
>
>
> On Thu, Aug 22, 2019 at 3:06 PM Curtis E. Combs Jr.
> wrote:
>>
>> Seems like the STP options are so common and necessary that it would
>> be a priority over seldom-used
On Thu, Aug 22, 2019 at 3:06 PM Curtis E. Combs Jr.
wrote:
> Seems like the STP options are so common and necessary that it would
> be a priority over seldom-used bridge_opts. I know what STP is and I'm
> not even a networking guy - never even heard of half of the
> bridge_opts that have
Seems like the STP options are so common and necessary that it would
be a priority over seldom-used bridge_opts. I know what STP is and I'm
not even a networking guy - never even heard of half of the
bridge_opts that have switches in the UI.
Anyway. I wanted to try the openvswitches, so I
Hi Dominik, would you mind sharing the use case for stp via API Only? I am
keen to know this.
Thanks
On Thu., 22 Aug. 2019, 19:24 Dominik Holler, wrote:
>
>
> On Thu, Aug 22, 2019 at 1:08 PM Miguel Duarte de Mora Barroso <
> mdbarr...@redhat.com> wrote:
>
>> On Sat, Aug 17, 2019 at 11:27 AM
On Thu, Aug 22, 2019 at 1:08 PM Miguel Duarte de Mora Barroso <
mdbarr...@redhat.com> wrote:
> On Sat, Aug 17, 2019 at 11:27 AM wrote:
> >
> > Hello. I have been trying to figure out an issue for a very long time.
> > That issue relates to the ethernet and 10gb fc links that I have on my
> >
On Sat, Aug 17, 2019 at 11:27 AM wrote:
>
> Hello. I have been trying to figure out an issue for a very long time.
> That issue relates to the ethernet and 10gb fc links that I have on my
> cluster being disabled any time a migration occurs.
>
> I believe this is because I need to have STP turned
I can see, just through a series of searches, that you can configure ovirt
to use a variety of things. I've configured many VMWare clusters before, so
having a (Cisco calls them) trunked port or (what other people call them)
ports in "tagged mode" would be nice, but like I was saying: I can't even
Oh, I'm just using the standard Linux bridges. I believe in ovirt
nomenclature they're called "legacy bridges". This is not by any real
choice but just what is created by default, I believe. In the first link
you sent, they are calling them "Linux legacy networking model" - though I
only have 2
No - no recommendations from me to use either. I took it that you were
using ovs bridge as I was not aware of another bridge. The only other
option I was aware of if what I am using, vlan interfaces and kernel vlan
tags. If you could share a link to what you're using, I would be keen to
read up on
Tony,
Well, thank you for that, but I'm not using openvswitchs, I'm just using
regular bridges. Are you suggesting that I do?
>From what I can see in the interface they have "(experimental)" marked on
them and we'd like to see production with this at some point.
None of my ports are trunking
A couple of links I found helpful, thought I'd send them over
http://therandomsecurityguy.com/openvswitch-cheat-sheet/
https://ovirt.org/develop/release-management/features/network/openvswitch/native-openvswitch.html
With STP off, if the network is detecting a loop then it will have to
block a
Cool, I can capture some packets tomorrow when I'm in the office and
see how that compares...
But, yea, it's a hassle to get them to respond IF they do, so the only
real options I'm going to have are what I can do with my servers from the OS. No
physical access. No nice DC guy to help me out.
If there is some other configuration that I'm not seeing, please do
let me know. I'd love to know more about what the expected
configuration is that enables VM migration and how it's "supposed" to
be configured. I have 3 hosts - each of these has two bridges. Each of
these bridges are named the
Hey Tony!
I only know the basics of Spanning Tree. At the current moment the
only way to get migrations to work at all without breaking the whole
oVirt cluster is to have it on. After changing it according to Paul's
instruction, it works like it has never worked before. Every migration
event was
e.albany,
STP is meant to block loops in layer 2. In basic operation, a root
bridge is elected which is the root of the tree. This bridge sends,
essentially 'hello' messages as multicast packets. The switches then
detect the loop in the network and block one of the links to prevent
such things as
I haven't used FC with oVirt but in the following it shows the bridge options
available and how to enable Ethtool and FCoE.
https://ovirt.org/documentation/admin-guide/appe-Custom_Network_Properties.html
Regards,
Paul S.
From:
on your physical
switch.
Regards,
Paul S.
From: Curtis E. Combs Jr.
Sent: 19 August 2019 12:39
To: Staniforth, Paul
Cc: users@ovirt.org
Subject: Re: [ovirt-users] Need to enable STP on ovirt bridges
Hi Paul,
Thank you very much for the response. However, I
Hi Paul,
Thank you very much for the response. However, I have tried those
options. Adding "stp=on" does not enable STP on the bridges. Each node
has a 1Gb Ethernet connection as well as a 10Gb FC connection, I've
tried moving the VM migration network to each and the effect is the
same.
The vdsm
39 matches
Mail list logo
