Re: [ovirt-users] iSCSI Multipathing -> host inactive

2016-08-29 Thread InterNetX - Juergen Gotteswinter
Am 29.08.2016 um 12:25 schrieb Nir Soffer:
> On Thu, Aug 25, 2016 at 2:37 PM, InterNetX - Juergen Gotteswinter
>  wrote:
>> currently, iscsi multipathed with solaris based filer as backend. but
>> this is already in progress of getting migrated to a different, less
>> fragile, plattform. ovirt is nice, but too bleeding edge and way to much
>> acting like a girly
> 
> "acting like a girly" is not  appropriate  for this list.
> 
> Nir
> 

i am sorry, this was never meant to discriminate any human. if it did, i
promise that it was not meant to.
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [ovirt-users] iSCSI Multipathing -> host inactive

2016-08-29 Thread Nir Soffer
On Thu, Aug 25, 2016 at 2:37 PM, InterNetX - Juergen Gotteswinter
 wrote:
> currently, iscsi multipathed with solaris based filer as backend. but
> this is already in progress of getting migrated to a different, less
> fragile, plattform. ovirt is nice, but too bleeding edge and way to much
> acting like a girly

"acting like a girly" is not  appropriate  for this list.

Nir
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [ovirt-users] iSCSI Multipathing -> host inactive

2016-08-29 Thread Nir Soffer
On Wed, Aug 24, 2016 at 6:15 PM, InterNetX - Juergen Gotteswinter
 wrote:
>
> iSCSI & Ovirt is an awful combination, no matter if multipathed or
> bonded. its always gambling how long it will work, and when it fails why
> did it fail.
>
> its supersensitive to latency,

Can you elaborate on this?

> and superfast with setting an host to
> inactive because the engine thinks something is wrong with it.

Typically it takes at least 5 minutes with abnormal monitoring
conditions before engine will make a host non-operational,
is this realy superfast?

> in most
> cases there was no real reason for.

I think the main issue was mixing of storage monitoring and lvm
refreshes, unneeded serialization of lvm commands, and bad
locking in engine side. The engine side was fixed in 3.6, and
the vdsm side in 4.0.
See https://bugzilla.redhat.com/1081962

In rhel/centos 7.2, lot of lot of multipath related issues were fixed,
and ovirt multipath configuration was fixed to prevent unwanted
io queuing with some devices, that could lead to long delays and
failures in many flows. However, I think our configuration is too
extreme, and you may like to use the configuration
in this patch:
https://gerrit.ovirt.org/61281

I guess trying 4.0 may be too bleeding edge for you, but
hopefully you will find that your iscsi setup is much more reliable
now.

Please file bugs if you still have issues with 4.0.

Nir

> we had this in several different hardware combinations, self built
> filers up on FreeBSD/Illumos & ZFS, Equallogic SAN, Nexenta Filer
>
> Been there, done that, wont do again.
>
> Am 24.08.2016 um 16:04 schrieb Uwe Laverenz:
> > Hi Elad,
> >
> > thank you very much for clearing things up.
> >
> > Initiator/iface 'a' tries to connect target 'b' and vice versa. As 'a'
> > and 'b' are in completely separate networks this can never work as long
> > as there is no routing between the networks.
> >
> > So it seems the iSCSI-bonding feature is not useful for my setup. I
> > still wonder how and where this feature is supposed to be used?
> >
> > thank you,
> > Uwe
> >
> > Am 24.08.2016 um 15:35 schrieb Elad Ben Aharon:
> >> Thanks.
> >>
> >> You're getting an iSCSI connection timeout [1], [2]. It means the host
> >> cannot connect to the targets from iface: enp9s0f1 nor iface: enp9s0f0.
> >>
> >> This causes the host to loose its connection to the storage and also,
> >> the connection to the engine becomes inactive. Therefore, the host
> >> changes its status to Non-responsive [3] and since it's the SPM, the
> >> whole DC, with all its storage domains become inactive.
> >>
> >>
> >> vdsm.log:
> >> [1]
> >> Traceback (most recent call last):
> >>   File "/usr/share/vdsm/storage/hsm.py", line 2400, in
> >> connectStorageServer
> >> conObj.connect()
> >>   File "/usr/share/vdsm/storage/storageServer.py", line 508, in connect
> >> iscsi.addIscsiNode(self._iface, self._target, self._cred)
> >>   File "/usr/share/vdsm/storage/iscsi.py", line 204, in addIscsiNode
> >> iscsiadm.node_login(iface.name , portalStr,
> >> target.iqn)
> >>   File "/usr/share/vdsm/storage/iscsiadm.py", line 336, in node_login
> >> raise IscsiNodeError(rc, out, err)
> >> IscsiNodeError: (8, ['Logging in to [iface: enp9s0f0, target:
> >> iqn.2005-10.org.freenas.ctl:tgtb, portal: 10.0.132.121,3260]
> >> (multiple)'], ['iscsiadm: Could not login to [iface: enp9s0f0, targ
> >> et: iqn.2005-10.org.freenas.ctl:tgtb, portal: 10.0.132.121,3260].',
> >> 'iscsiadm: initiator reported error (8 - connection timed out)',
> >> 'iscsiadm: Could not log into all portals'])
> >>
> >>
> >>
> >> vdsm.log:
> >> [2]
> >> Traceback (most recent call last):
> >>   File "/usr/share/vdsm/storage/hsm.py", line 2400, in
> >> connectStorageServer
> >> conObj.connect()
> >>   File "/usr/share/vdsm/storage/storageServer.py", line 508, in connect
> >> iscsi.addIscsiNode(self._iface, self._target, self._cred)
> >>   File "/usr/share/vdsm/storage/iscsi.py", line 204, in addIscsiNode
> >> iscsiadm.node_login(iface.name , portalStr,
> >> target.iqn)
> >>   File "/usr/share/vdsm/storage/iscsiadm.py", line 336, in node_login
> >> raise IscsiNodeError(rc, out, err)
> >> IscsiNodeError: (8, ['Logging in to [iface: enp9s0f1, target:
> >> iqn.2005-10.org.freenas.ctl:tgta, portal: 10.0.131.121,3260]
> >> (multiple)'], ['iscsiadm: Could not login to [iface: enp9s0f1, target:
> >> iqn.2005-10.org.freenas.ctl:tgta, portal: 10.0.131.121,3260].',
> >> 'iscsiadm: initiator reported error (8 - connection timed out)',
> >> 'iscsiadm: Could not log into all portals'])
> >>
> >>
> >> engine.log:
> >> [3]
> >>
> >>
> >> 2016-08-24 14:10:23,222 WARN
> >> [org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector]
> >> (default task-25) [15d1637f] Correlation ID: 15d1637f, Call Stack: null,
> >> Custom Event ID:
> >>  -1, Message: iSCSI bond 'iBond' was successfully created in Data Center
> >> 'Default' 

Re: [ovirt-users] iSCSI Multipathing -> host inactive

2016-08-26 Thread Yaniv Kaul
On Fri, Aug 26, 2016 at 1:33 PM, InterNetX - Juergen Gotteswinter <
j...@internetx.com> wrote:

>
>
> Am 25.08.2016 um 15:53 schrieb Yaniv Kaul:
> >
> >
> > On Wed, Aug 24, 2016 at 6:15 PM, InterNetX - Juergen Gotteswinter
> >  > > wrote:
> >
> > iSCSI & Ovirt is an awful combination, no matter if multipathed or
> > bonded. its always gambling how long it will work, and when it fails
> why
> > did it fail.
> >
> >
> > I disagree. In most cases, it's actually a lower layer issues. In most
> > cases, btw, it's because multipathing was not configured (or not
> > configured correctly).
> >
>
> experience tells me it is like is said, this was something i have seen
> from 3.0 up to 3.6. Ovirt, and - suprprise - RHEV. Both act the same
> way. I am absolutly aware of Multpath Configurations, iSCSI Multipathing
> is very widespread in use in our DC. But such problems are an excluse
> Ovirt/RHEV Feature.
>

I don't think the resentful tone is appropriate for the oVirt community
mailing list.


>
> >
> >
> > its supersensitive to latency, and superfast with setting an host to
> > inactive because the engine thinks something is wrong with it. in
> most
> > cases there was no real reason for.
> >
> >
> > Did you open bugs for those issues? I'm not aware of 'no real reason'
> > issues.
> >
>
> Support Tickets for Rhev Installation, after Support (even after massive
> escalation requests) kept telling me the same again and again i gave up
> and we dropped the RHEV Subscriptions to Migrate the VMS to a different
> Plattform Solution (still iSCSI Backend). Problems gone.
>

I wish you peace of mind with your new platform solution.
>From a (shallow!) search I've made on oVirt bugs, I could not find any
oVirt issue you've reported or commented on.
I am aware of the request to set rp_filter correctly for setups with
multiple interfaces in the same IP subnet.


>
>
> >
> >
> > we had this in several different hardware combinations, self built
> > filers up on FreeBSD/Illumos & ZFS, Equallogic SAN, Nexenta Filer
> >
> > Been there, done that, wont do again.
> >
> >
> > We've had good success and reliability with most enterprise level
> > storage, such as EMC, NetApp, Dell filers.
> > When properly configured, of course.
> > Y.
> >
>
> Dell Equallogic? Cant really believe since Ovirt / Rhev and the
> Equallogic Network Configuration wont play nice together (EQL wants all
> Interfaces in the same Subnet). And they only work like expected when
> there Hit Kit Driverpackage is installed. Without Path Failover is like
> russian Roulette. But Ovirt hates the Hit Kit, so this Combo ends up in
> a huge mess, because Ovirt does changes to iSCSI, as well as the Hit Kit
> -> Kaboom. Host not available.
>

Thanks - I'll look into this specific storage.
I'm aware it's unique in some cases, but I don't have experience with it
specifically.


>
> There are several KB Articles in the RHN, without real Solution.
>
>
> But like you try to tell between the lines, this must be the Customers
> misconfiguration. Yep, typical Supportkilleranswer. Same Style than in
> RHN Tickets, i am done with this.
>

A funny sentence I've read yesterday:
Schrodinger's backup: "the condition of any backup is unknown until restore
is attempted."

In a sense, this is similar to no SPOF high availability setup - in many
cases you don't know it works well until needed.
There are simply many variables and components involved.
That was all I meant, nothing between the lines and I apologize if I've
given you a different impression.
Y.


> Thanks.
>
> >
> >
> >
> > Am 24.08.2016 um 16:04 schrieb Uwe Laverenz:
> > > Hi Elad,
> > >
> > > thank you very much for clearing things up.
> > >
> > > Initiator/iface 'a' tries to connect target 'b' and vice versa. As
> 'a'
> > > and 'b' are in completely separate networks this can never work as
> > long
> > > as there is no routing between the networks.
> > >
> > > So it seems the iSCSI-bonding feature is not useful for my setup. I
> > > still wonder how and where this feature is supposed to be used?
> > >
> > > thank you,
> > > Uwe
> > >
> > > Am 24.08.2016 um 15:35 schrieb Elad Ben Aharon:
> > >> Thanks.
> > >>
> > >> You're getting an iSCSI connection timeout [1], [2]. It means the
> > host
> > >> cannot connect to the targets from iface: enp9s0f1 nor iface:
> > enp9s0f0.
> > >>
> > >> This causes the host to loose its connection to the storage and
> also,
> > >> the connection to the engine becomes inactive. Therefore, the host
> > >> changes its status to Non-responsive [3] and since it's the SPM,
> the
> > >> whole DC, with all its storage domains become inactive.
> > >>
> > >>
> > >> vdsm.log:
> > >> [1]
> > >> Traceback (most recent call last):
> > >>   File 

Re: [ovirt-users] iSCSI Multipathing -> host inactive

2016-08-26 Thread Nicolas Ecarnot

Le 26/08/2016 à 12:33, InterNetX - Juergen Gotteswinter a écrit :



Am 25.08.2016 um 15:53 schrieb Yaniv Kaul:



On Wed, Aug 24, 2016 at 6:15 PM, InterNetX - Juergen Gotteswinter
> wrote:

iSCSI & Ovirt is an awful combination, no matter if multipathed or
bonded. its always gambling how long it will work, and when it fails why
did it fail.


I disagree. In most cases, it's actually a lower layer issues. In most
cases, btw, it's because multipathing was not configured (or not
configured correctly).



experience tells me it is like is said, this was something i have seen
from 3.0 up to 3.6. Ovirt, and - suprprise - RHEV. Both act the same
way. I am absolutly aware of Multpath Configurations, iSCSI Multipathing
is very widespread in use in our DC. But such problems are an excluse
Ovirt/RHEV Feature.




its supersensitive to latency, and superfast with setting an host to
inactive because the engine thinks something is wrong with it. in most
cases there was no real reason for.


Did you open bugs for those issues? I'm not aware of 'no real reason'
issues.



Support Tickets for Rhev Installation, after Support (even after massive
escalation requests) kept telling me the same again and again i gave up
and we dropped the RHEV Subscriptions to Migrate the VMS to a different
Plattform Solution (still iSCSI Backend). Problems gone.





we had this in several different hardware combinations, self built
filers up on FreeBSD/Illumos & ZFS, Equallogic SAN, Nexenta Filer

Been there, done that, wont do again.


We've had good success and reliability with most enterprise level
storage, such as EMC, NetApp, Dell filers.
When properly configured, of course.
Y.



Dell Equallogic? Cant really believe since Ovirt / Rhev and the
Equallogic Network Configuration wont play nice together (EQL wants all
Interfaces in the same Subnet). And they only work like expected when
there Hit Kit Driverpackage is installed. Without Path Failover is like
russian Roulette. But Ovirt hates the Hit Kit, so this Combo ends up in
a huge mess, because Ovirt does changes to iSCSI, as well as the Hit Kit
-> Kaboom. Host not available.

There are several KB Articles in the RHN, without real Solution.


I'm working with Dell, EMC and EQL for 7 years, and with oVirt for 4 
years, and I must recognize that what Juergen said is true : HIT is not 
compatible with the way oVirt is using iSCSI (or the other way round).

What he says about LVM is also true.
I still love oVirt, and won't quit. I'm just very patient and hoping my 
BZ and RFE get fixed.



--
Nicolas ECARNOT
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [ovirt-users] iSCSI Multipathing -> host inactive

2016-08-26 Thread InterNetX - Juergen Gotteswinter
one more thing, which i am sure that most ppl are not aware of.

when using thin provisioned disks for vms, hosted on iSCSI SAN Ovirt
uses a for me unusual way to do this.

Ovirt adds a new LVM LV for a vm,  generates a Thin Qcow Image which is
written directly raw onto that LV. So for, ok, can be done like this.

But try generating some Write Load from within the Guest and see
yourself what will happen. Supports answer to this is: use Raw, without
thin.

Seems to me like a wrong design decision, and i can only warn everyone
to use this.

Am 26.08.2016 um 12:33 schrieb InterNetX - Juergen Gotteswinter:
> 
> 
> Am 25.08.2016 um 15:53 schrieb Yaniv Kaul:
>>
>>
>> On Wed, Aug 24, 2016 at 6:15 PM, InterNetX - Juergen Gotteswinter
>> > > wrote:
>>
>> iSCSI & Ovirt is an awful combination, no matter if multipathed or
>> bonded. its always gambling how long it will work, and when it fails why
>> did it fail.
>>
>>
>> I disagree. In most cases, it's actually a lower layer issues. In most
>> cases, btw, it's because multipathing was not configured (or not
>> configured correctly).
>>  
> 
> experience tells me it is like is said, this was something i have seen
> from 3.0 up to 3.6. Ovirt, and - suprprise - RHEV. Both act the same
> way. I am absolutly aware of Multpath Configurations, iSCSI Multipathing
> is very widespread in use in our DC. But such problems are an excluse
> Ovirt/RHEV Feature.
> 
>>
>>
>> its supersensitive to latency, and superfast with setting an host to
>> inactive because the engine thinks something is wrong with it. in most
>> cases there was no real reason for.
>>
>>
>> Did you open bugs for those issues? I'm not aware of 'no real reason'
>> issues.
>>  
> 
> Support Tickets for Rhev Installation, after Support (even after massive
> escalation requests) kept telling me the same again and again i gave up
> and we dropped the RHEV Subscriptions to Migrate the VMS to a different
> Plattform Solution (still iSCSI Backend). Problems gone.
> 
> 
>>
>>
>> we had this in several different hardware combinations, self built
>> filers up on FreeBSD/Illumos & ZFS, Equallogic SAN, Nexenta Filer
>>
>> Been there, done that, wont do again.
>>
>>
>> We've had good success and reliability with most enterprise level
>> storage, such as EMC, NetApp, Dell filers.
>> When properly configured, of course.
>> Y.
>>
> 
> Dell Equallogic? Cant really believe since Ovirt / Rhev and the
> Equallogic Network Configuration wont play nice together (EQL wants all
> Interfaces in the same Subnet). And they only work like expected when
> there Hit Kit Driverpackage is installed. Without Path Failover is like
> russian Roulette. But Ovirt hates the Hit Kit, so this Combo ends up in
> a huge mess, because Ovirt does changes to iSCSI, as well as the Hit Kit
> -> Kaboom. Host not available.
> 
> There are several KB Articles in the RHN, without real Solution.
> 
> 
> But like you try to tell between the lines, this must be the Customers
> misconfiguration. Yep, typical Supportkilleranswer. Same Style than in
> RHN Tickets, i am done with this.
> 
> Thanks.
> 
>>  
>>
>>
>> Am 24.08.2016 um 16:04 schrieb Uwe Laverenz:
>> > Hi Elad,
>> >
>> > thank you very much for clearing things up.
>> >
>> > Initiator/iface 'a' tries to connect target 'b' and vice versa. As 'a'
>> > and 'b' are in completely separate networks this can never work as
>> long
>> > as there is no routing between the networks.
>> >
>> > So it seems the iSCSI-bonding feature is not useful for my setup. I
>> > still wonder how and where this feature is supposed to be used?
>> >
>> > thank you,
>> > Uwe
>> >
>> > Am 24.08.2016 um 15:35 schrieb Elad Ben Aharon:
>> >> Thanks.
>> >>
>> >> You're getting an iSCSI connection timeout [1], [2]. It means the
>> host
>> >> cannot connect to the targets from iface: enp9s0f1 nor iface:
>> enp9s0f0.
>> >>
>> >> This causes the host to loose its connection to the storage and also,
>> >> the connection to the engine becomes inactive. Therefore, the host
>> >> changes its status to Non-responsive [3] and since it's the SPM, the
>> >> whole DC, with all its storage domains become inactive.
>> >>
>> >>
>> >> vdsm.log:
>> >> [1]
>> >> Traceback (most recent call last):
>> >>   File "/usr/share/vdsm/storage/hsm.py", line 2400, in
>> >> connectStorageServer
>> >> conObj.connect()
>> >>   File "/usr/share/vdsm/storage/storageServer.py", line 508, in
>> connect
>> >> iscsi.addIscsiNode(self._iface, self._target, self._cred)
>> >>   File "/usr/share/vdsm/storage/iscsi.py", line 204, in addIscsiNode
>> >> iscsiadm.node_login(iface.name 
>> , portalStr,
>> >> target.iqn)
>> >>   File 

Re: [ovirt-users] iSCSI Multipathing -> host inactive

2016-08-25 Thread Yaniv Kaul
On Wed, Aug 24, 2016 at 6:15 PM, InterNetX - Juergen Gotteswinter <
juergen.gotteswin...@internetx.com> wrote:

> iSCSI & Ovirt is an awful combination, no matter if multipathed or
> bonded. its always gambling how long it will work, and when it fails why
> did it fail.
>

I disagree. In most cases, it's actually a lower layer issues. In most
cases, btw, it's because multipathing was not configured (or not configured
correctly).


>
> its supersensitive to latency, and superfast with setting an host to
> inactive because the engine thinks something is wrong with it. in most
> cases there was no real reason for.
>

Did you open bugs for those issues? I'm not aware of 'no real reason'
issues.


>
> we had this in several different hardware combinations, self built
> filers up on FreeBSD/Illumos & ZFS, Equallogic SAN, Nexenta Filer
>
> Been there, done that, wont do again.
>

We've had good success and reliability with most enterprise level storage,
such as EMC, NetApp, Dell filers.
When properly configured, of course.
Y.



>
> Am 24.08.2016 um 16:04 schrieb Uwe Laverenz:
> > Hi Elad,
> >
> > thank you very much for clearing things up.
> >
> > Initiator/iface 'a' tries to connect target 'b' and vice versa. As 'a'
> > and 'b' are in completely separate networks this can never work as long
> > as there is no routing between the networks.
> >
> > So it seems the iSCSI-bonding feature is not useful for my setup. I
> > still wonder how and where this feature is supposed to be used?
> >
> > thank you,
> > Uwe
> >
> > Am 24.08.2016 um 15:35 schrieb Elad Ben Aharon:
> >> Thanks.
> >>
> >> You're getting an iSCSI connection timeout [1], [2]. It means the host
> >> cannot connect to the targets from iface: enp9s0f1 nor iface: enp9s0f0.
> >>
> >> This causes the host to loose its connection to the storage and also,
> >> the connection to the engine becomes inactive. Therefore, the host
> >> changes its status to Non-responsive [3] and since it's the SPM, the
> >> whole DC, with all its storage domains become inactive.
> >>
> >>
> >> vdsm.log:
> >> [1]
> >> Traceback (most recent call last):
> >>   File "/usr/share/vdsm/storage/hsm.py", line 2400, in
> >> connectStorageServer
> >> conObj.connect()
> >>   File "/usr/share/vdsm/storage/storageServer.py", line 508, in connect
> >> iscsi.addIscsiNode(self._iface, self._target, self._cred)
> >>   File "/usr/share/vdsm/storage/iscsi.py", line 204, in addIscsiNode
> >> iscsiadm.node_login(iface.name , portalStr,
> >> target.iqn)
> >>   File "/usr/share/vdsm/storage/iscsiadm.py", line 336, in node_login
> >> raise IscsiNodeError(rc, out, err)
> >> IscsiNodeError: (8, ['Logging in to [iface: enp9s0f0, target:
> >> iqn.2005-10.org.freenas.ctl:tgtb, portal: 10.0.132.121,3260]
> >> (multiple)'], ['iscsiadm: Could not login to [iface: enp9s0f0, targ
> >> et: iqn.2005-10.org.freenas.ctl:tgtb, portal: 10.0.132.121,3260].',
> >> 'iscsiadm: initiator reported error (8 - connection timed out)',
> >> 'iscsiadm: Could not log into all portals'])
> >>
> >>
> >>
> >> vdsm.log:
> >> [2]
> >> Traceback (most recent call last):
> >>   File "/usr/share/vdsm/storage/hsm.py", line 2400, in
> >> connectStorageServer
> >> conObj.connect()
> >>   File "/usr/share/vdsm/storage/storageServer.py", line 508, in connect
> >> iscsi.addIscsiNode(self._iface, self._target, self._cred)
> >>   File "/usr/share/vdsm/storage/iscsi.py", line 204, in addIscsiNode
> >> iscsiadm.node_login(iface.name , portalStr,
> >> target.iqn)
> >>   File "/usr/share/vdsm/storage/iscsiadm.py", line 336, in node_login
> >> raise IscsiNodeError(rc, out, err)
> >> IscsiNodeError: (8, ['Logging in to [iface: enp9s0f1, target:
> >> iqn.2005-10.org.freenas.ctl:tgta, portal: 10.0.131.121,3260]
> >> (multiple)'], ['iscsiadm: Could not login to [iface: enp9s0f1, target:
> >> iqn.2005-10.org.freenas.ctl:tgta, portal: 10.0.131.121,3260].',
> >> 'iscsiadm: initiator reported error (8 - connection timed out)',
> >> 'iscsiadm: Could not log into all portals'])
> >>
> >>
> >> engine.log:
> >> [3]
> >>
> >>
> >> 2016-08-24 14:10:23,222 WARN
> >> [org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector]
> >> (default task-25) [15d1637f] Correlation ID: 15d1637f, Call Stack: null,
> >> Custom Event ID:
> >>  -1, Message: iSCSI bond 'iBond' was successfully created in Data Center
> >> 'Default' but some of the hosts encountered connection issues.
> >>
> >>
> >>
> >> 2016-08-24 14:10:23,208 INFO
> >> [org.ovirt.engine.core.vdsbroker.vdsbroker.
> ConnectStorageServerVDSCommand]
> >>
> >> (org.ovirt.thread.pool-8-thread-25) [15d1637f] Command
> >> 'org.ovirt.engine.core.vdsbrok
> >> er.vdsbroker.ConnectStorageServerVDSCommand' return value '
> >> ServerConnectionStatusReturnForXmlRpc:{status='StatusForXmlRpc
> >> [code=5022, message=Message timeout which can be caused by communication
> >> issues]'}
> >>
> >>
> >>
> >> On Wed, Aug 24, 2016 at 4:04 PM, Uwe Laverenz 

Re: [ovirt-users] iSCSI Multipathing -> host inactive

2016-08-25 Thread Nicolas Ecarnot

Le 25/08/2016 à 13:37, InterNetX - Juergen Gotteswinter a écrit :



Am 24.08.2016 um 17:15 schrieb InterNetX - Juergen Gotteswinter:

iSCSI & Ovirt is an awful combination, no matter if multipathed or
bonded. its always gambling how long it will work, and when it fails why
did it fail.


We are using oVirt + iSCSI, but using bonding mode 1 (active/passive) 
and no multipathing.


It is running OK. So far.



its supersensitive to latency, and superfast with setting an host to
inactive because the engine thinks something is wrong with it. in most
cases there was no real reason for.


You are not wrong...



we had this in several different hardware combinations, self built
filers up on FreeBSD/Illumos & ZFS, Equallogic SAN, Nexenta Filer

Been there, done that, wont do again.


Thank you, I take this as a warning. :)

For my testbed I chose to ignore the iSCSI-bond feature and change the
multipath default to round robin instead.

What kind of storage do you use in production? Fibre channel, gluster,
ceph, ...?




we had this in several different hardware combinations, self built
filers up on FreeBSD/Illumos & ZFS, Equallogic SAN, Nexenta Filer


currently, iscsi multipathed with solaris based filer as backend. but
this is already in progress of getting migrated to a different, less
fragile, plattform. ovirt is nice, but too bleeding edge and way to much




acting like a _girly_


Man, you killed me! :) :) :)

--
Nicolas ECARNOT, crying laughing :) :) :)
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [ovirt-users] iSCSI Multipathing -> host inactive

2016-08-25 Thread InterNetX - Juergen Gotteswinter


Am 25.08.2016 um 08:42 schrieb Uwe Laverenz:
> Hi Jürgen,
> 
> Am 24.08.2016 um 17:15 schrieb InterNetX - Juergen Gotteswinter:
>> iSCSI & Ovirt is an awful combination, no matter if multipathed or
>> bonded. its always gambling how long it will work, and when it fails why
>> did it fail.
>>
>> its supersensitive to latency, and superfast with setting an host to
>> inactive because the engine thinks something is wrong with it. in most
>> cases there was no real reason for.
>>
>> we had this in several different hardware combinations, self built
>> filers up on FreeBSD/Illumos & ZFS, Equallogic SAN, Nexenta Filer
>>
>> Been there, done that, wont do again.
> 
> Thank you, I take this as a warning. :)
> 
> For my testbed I chose to ignore the iSCSI-bond feature and change the
> multipath default to round robin instead.
> 
> What kind of storage do you use in production? Fibre channel, gluster,
> ceph, ...?

>>
>> we had this in several different hardware combinations, self built
>> filers up on FreeBSD/Illumos & ZFS, Equallogic SAN, Nexenta Filer

currently, iscsi multipathed with solaris based filer as backend. but
this is already in progress of getting migrated to a different, less
fragile, plattform. ovirt is nice, but too bleeding edge and way to much
acting like a girly

> 
> thanks,
> Uwe
> ___
> Users mailing list
> Users@ovirt.org
> http://lists.ovirt.org/mailman/listinfo/users
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [ovirt-users] iSCSI Multipathing -> host inactive

2016-08-25 Thread Uwe Laverenz

Hi Jürgen,

Am 24.08.2016 um 17:15 schrieb InterNetX - Juergen Gotteswinter:

iSCSI & Ovirt is an awful combination, no matter if multipathed or
bonded. its always gambling how long it will work, and when it fails why
did it fail.

its supersensitive to latency, and superfast with setting an host to
inactive because the engine thinks something is wrong with it. in most
cases there was no real reason for.

we had this in several different hardware combinations, self built
filers up on FreeBSD/Illumos & ZFS, Equallogic SAN, Nexenta Filer

Been there, done that, wont do again.


Thank you, I take this as a warning. :)

For my testbed I chose to ignore the iSCSI-bond feature and change the 
multipath default to round robin instead.


What kind of storage do you use in production? Fibre channel, gluster, 
ceph, ...?


thanks,
Uwe
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [ovirt-users] iSCSI Multipathing -> host inactive

2016-08-24 Thread InterNetX - Juergen Gotteswinter
iSCSI & Ovirt is an awful combination, no matter if multipathed or
bonded. its always gambling how long it will work, and when it fails why
did it fail.

its supersensitive to latency, and superfast with setting an host to
inactive because the engine thinks something is wrong with it. in most
cases there was no real reason for.

we had this in several different hardware combinations, self built
filers up on FreeBSD/Illumos & ZFS, Equallogic SAN, Nexenta Filer

Been there, done that, wont do again.

Am 24.08.2016 um 16:04 schrieb Uwe Laverenz:
> Hi Elad,
> 
> thank you very much for clearing things up.
> 
> Initiator/iface 'a' tries to connect target 'b' and vice versa. As 'a'
> and 'b' are in completely separate networks this can never work as long
> as there is no routing between the networks.
> 
> So it seems the iSCSI-bonding feature is not useful for my setup. I
> still wonder how and where this feature is supposed to be used?
> 
> thank you,
> Uwe
> 
> Am 24.08.2016 um 15:35 schrieb Elad Ben Aharon:
>> Thanks.
>>
>> You're getting an iSCSI connection timeout [1], [2]. It means the host
>> cannot connect to the targets from iface: enp9s0f1 nor iface: enp9s0f0.
>>
>> This causes the host to loose its connection to the storage and also,
>> the connection to the engine becomes inactive. Therefore, the host
>> changes its status to Non-responsive [3] and since it's the SPM, the
>> whole DC, with all its storage domains become inactive.
>>
>>
>> vdsm.log:
>> [1]
>> Traceback (most recent call last):
>>   File "/usr/share/vdsm/storage/hsm.py", line 2400, in
>> connectStorageServer
>> conObj.connect()
>>   File "/usr/share/vdsm/storage/storageServer.py", line 508, in connect
>> iscsi.addIscsiNode(self._iface, self._target, self._cred)
>>   File "/usr/share/vdsm/storage/iscsi.py", line 204, in addIscsiNode
>> iscsiadm.node_login(iface.name , portalStr,
>> target.iqn)
>>   File "/usr/share/vdsm/storage/iscsiadm.py", line 336, in node_login
>> raise IscsiNodeError(rc, out, err)
>> IscsiNodeError: (8, ['Logging in to [iface: enp9s0f0, target:
>> iqn.2005-10.org.freenas.ctl:tgtb, portal: 10.0.132.121,3260]
>> (multiple)'], ['iscsiadm: Could not login to [iface: enp9s0f0, targ
>> et: iqn.2005-10.org.freenas.ctl:tgtb, portal: 10.0.132.121,3260].',
>> 'iscsiadm: initiator reported error (8 - connection timed out)',
>> 'iscsiadm: Could not log into all portals'])
>>
>>
>>
>> vdsm.log:
>> [2]
>> Traceback (most recent call last):
>>   File "/usr/share/vdsm/storage/hsm.py", line 2400, in
>> connectStorageServer
>> conObj.connect()
>>   File "/usr/share/vdsm/storage/storageServer.py", line 508, in connect
>> iscsi.addIscsiNode(self._iface, self._target, self._cred)
>>   File "/usr/share/vdsm/storage/iscsi.py", line 204, in addIscsiNode
>> iscsiadm.node_login(iface.name , portalStr,
>> target.iqn)
>>   File "/usr/share/vdsm/storage/iscsiadm.py", line 336, in node_login
>> raise IscsiNodeError(rc, out, err)
>> IscsiNodeError: (8, ['Logging in to [iface: enp9s0f1, target:
>> iqn.2005-10.org.freenas.ctl:tgta, portal: 10.0.131.121,3260]
>> (multiple)'], ['iscsiadm: Could not login to [iface: enp9s0f1, target:
>> iqn.2005-10.org.freenas.ctl:tgta, portal: 10.0.131.121,3260].',
>> 'iscsiadm: initiator reported error (8 - connection timed out)',
>> 'iscsiadm: Could not log into all portals'])
>>
>>
>> engine.log:
>> [3]
>>
>>
>> 2016-08-24 14:10:23,222 WARN
>> [org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector]
>> (default task-25) [15d1637f] Correlation ID: 15d1637f, Call Stack: null,
>> Custom Event ID:
>>  -1, Message: iSCSI bond 'iBond' was successfully created in Data Center
>> 'Default' but some of the hosts encountered connection issues.
>>
>>
>>
>> 2016-08-24 14:10:23,208 INFO
>> [org.ovirt.engine.core.vdsbroker.vdsbroker.ConnectStorageServerVDSCommand]
>>
>> (org.ovirt.thread.pool-8-thread-25) [15d1637f] Command
>> 'org.ovirt.engine.core.vdsbrok
>> er.vdsbroker.ConnectStorageServerVDSCommand' return value '
>> ServerConnectionStatusReturnForXmlRpc:{status='StatusForXmlRpc
>> [code=5022, message=Message timeout which can be caused by communication
>> issues]'}
>>
>>
>>
>> On Wed, Aug 24, 2016 at 4:04 PM, Uwe Laverenz > > wrote:
>>
>> Hi Elad,
>>
>> I sent you a download message.
>>
>> thank you,
>> Uwe
>> ___
>> Users mailing list
>> Users@ovirt.org 
>> http://lists.ovirt.org/mailman/listinfo/users
>> 
>>
>>
> ___
> Users mailing list
> Users@ovirt.org
> http://lists.ovirt.org/mailman/listinfo/users
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [ovirt-users] iSCSI Multipathing -> host inactive

2016-08-24 Thread Uwe Laverenz

Hi Elad,

thank you very much for clearing things up.

Initiator/iface 'a' tries to connect target 'b' and vice versa. As 'a' 
and 'b' are in completely separate networks this can never work as long 
as there is no routing between the networks.


So it seems the iSCSI-bonding feature is not useful for my setup. I 
still wonder how and where this feature is supposed to be used?


thank you,
Uwe

Am 24.08.2016 um 15:35 schrieb Elad Ben Aharon:

Thanks.

You're getting an iSCSI connection timeout [1], [2]. It means the host
cannot connect to the targets from iface: enp9s0f1 nor iface: enp9s0f0.

This causes the host to loose its connection to the storage and also,
the connection to the engine becomes inactive. Therefore, the host
changes its status to Non-responsive [3] and since it's the SPM, the
whole DC, with all its storage domains become inactive.


vdsm.log:
[1]
Traceback (most recent call last):
  File "/usr/share/vdsm/storage/hsm.py", line 2400, in connectStorageServer
conObj.connect()
  File "/usr/share/vdsm/storage/storageServer.py", line 508, in connect
iscsi.addIscsiNode(self._iface, self._target, self._cred)
  File "/usr/share/vdsm/storage/iscsi.py", line 204, in addIscsiNode
iscsiadm.node_login(iface.name , portalStr,
target.iqn)
  File "/usr/share/vdsm/storage/iscsiadm.py", line 336, in node_login
raise IscsiNodeError(rc, out, err)
IscsiNodeError: (8, ['Logging in to [iface: enp9s0f0, target:
iqn.2005-10.org.freenas.ctl:tgtb, portal: 10.0.132.121,3260]
(multiple)'], ['iscsiadm: Could not login to [iface: enp9s0f0, targ
et: iqn.2005-10.org.freenas.ctl:tgtb, portal: 10.0.132.121,3260].',
'iscsiadm: initiator reported error (8 - connection timed out)',
'iscsiadm: Could not log into all portals'])



vdsm.log:
[2]
Traceback (most recent call last):
  File "/usr/share/vdsm/storage/hsm.py", line 2400, in connectStorageServer
conObj.connect()
  File "/usr/share/vdsm/storage/storageServer.py", line 508, in connect
iscsi.addIscsiNode(self._iface, self._target, self._cred)
  File "/usr/share/vdsm/storage/iscsi.py", line 204, in addIscsiNode
iscsiadm.node_login(iface.name , portalStr,
target.iqn)
  File "/usr/share/vdsm/storage/iscsiadm.py", line 336, in node_login
raise IscsiNodeError(rc, out, err)
IscsiNodeError: (8, ['Logging in to [iface: enp9s0f1, target:
iqn.2005-10.org.freenas.ctl:tgta, portal: 10.0.131.121,3260]
(multiple)'], ['iscsiadm: Could not login to [iface: enp9s0f1, target:
iqn.2005-10.org.freenas.ctl:tgta, portal: 10.0.131.121,3260].',
'iscsiadm: initiator reported error (8 - connection timed out)',
'iscsiadm: Could not log into all portals'])


engine.log:
[3]


2016-08-24 14:10:23,222 WARN
[org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector]
(default task-25) [15d1637f] Correlation ID: 15d1637f, Call Stack: null,
Custom Event ID:
 -1, Message: iSCSI bond 'iBond' was successfully created in Data Center
'Default' but some of the hosts encountered connection issues.



2016-08-24 14:10:23,208 INFO
[org.ovirt.engine.core.vdsbroker.vdsbroker.ConnectStorageServerVDSCommand]
(org.ovirt.thread.pool-8-thread-25) [15d1637f] Command
'org.ovirt.engine.core.vdsbrok
er.vdsbroker.ConnectStorageServerVDSCommand' return value '
ServerConnectionStatusReturnForXmlRpc:{status='StatusForXmlRpc
[code=5022, message=Message timeout which can be caused by communication
issues]'}



On Wed, Aug 24, 2016 at 4:04 PM, Uwe Laverenz > wrote:

Hi Elad,

I sent you a download message.

thank you,
Uwe
___
Users mailing list
Users@ovirt.org 
http://lists.ovirt.org/mailman/listinfo/users




___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [ovirt-users] iSCSI Multipathing -> host inactive

2016-08-24 Thread Elad Ben Aharon
Thanks.

You're getting an iSCSI connection timeout [1], [2]. It means the host
cannot connect to the targets from iface: enp9s0f1 nor iface: enp9s0f0.

This causes the host to loose its connection to the storage and also, the
connection to the engine becomes inactive. Therefore, the host changes its
status to Non-responsive [3] and since it's the SPM, the whole DC, with all
its storage domains become inactive.


vdsm.log:
[1]
Traceback (most recent call last):
  File "/usr/share/vdsm/storage/hsm.py", line 2400, in connectStorageServer
conObj.connect()
  File "/usr/share/vdsm/storage/storageServer.py", line 508, in connect
iscsi.addIscsiNode(self._iface, self._target, self._cred)
  File "/usr/share/vdsm/storage/iscsi.py", line 204, in addIscsiNode
iscsiadm.node_login(iface.name, portalStr, target.iqn)
  File "/usr/share/vdsm/storage/iscsiadm.py", line 336, in node_login
raise IscsiNodeError(rc, out, err)
IscsiNodeError: (8, ['Logging in to [iface: enp9s0f0, target:
iqn.2005-10.org.freenas.ctl:tgtb, portal: 10.0.132.121,3260] (multiple)'],
['iscsiadm: Could not login to [iface: enp9s0f0, targ
et: iqn.2005-10.org.freenas.ctl:tgtb, portal: 10.0.132.121,3260].',
'iscsiadm: initiator reported error (8 - connection timed out)', 'iscsiadm:
Could not log into all portals'])



vdsm.log:
[2]
Traceback (most recent call last):
  File "/usr/share/vdsm/storage/hsm.py", line 2400, in connectStorageServer
conObj.connect()
  File "/usr/share/vdsm/storage/storageServer.py", line 508, in connect
iscsi.addIscsiNode(self._iface, self._target, self._cred)
  File "/usr/share/vdsm/storage/iscsi.py", line 204, in addIscsiNode
iscsiadm.node_login(iface.name, portalStr, target.iqn)
  File "/usr/share/vdsm/storage/iscsiadm.py", line 336, in node_login
raise IscsiNodeError(rc, out, err)
IscsiNodeError: (8, ['Logging in to [iface: enp9s0f1, target:
iqn.2005-10.org.freenas.ctl:tgta, portal: 10.0.131.121,3260] (multiple)'],
['iscsiadm: Could not login to [iface: enp9s0f1, target:
iqn.2005-10.org.freenas.ctl:tgta, portal: 10.0.131.121,3260].', 'iscsiadm:
initiator reported error (8 - connection timed out)', 'iscsiadm: Could not
log into all portals'])


engine.log:
[3]


2016-08-24 14:10:23,222 WARN
[org.ovirt.engine.core.dal.dbbroker.auditloghandling.AuditLogDirector]
(default task-25) [15d1637f] Correlation ID: 15d1637f, Call Stack: null,
Custom Event ID:
 -1, Message: iSCSI bond 'iBond' was successfully created in Data Center
'Default' but some of the hosts encountered connection issues.



2016-08-24 14:10:23,208 INFO
[org.ovirt.engine.core.vdsbroker.vdsbroker.ConnectStorageServerVDSCommand]
(org.ovirt.thread.pool-8-thread-25) [15d1637f] Command
'org.ovirt.engine.core.vdsbrok
er.vdsbroker.ConnectStorageServerVDSCommand' return value '
ServerConnectionStatusReturnForXmlRpc:{status='StatusForXmlRpc [code=5022,
message=Message timeout which can be caused by communication issues]'}



On Wed, Aug 24, 2016 at 4:04 PM, Uwe Laverenz  wrote:

> Hi Elad,
>
> I sent you a download message.
>
> thank you,
> Uwe
> ___
> Users mailing list
> Users@ovirt.org
> http://lists.ovirt.org/mailman/listinfo/users
>
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [ovirt-users] iSCSI Multipathing -> host inactive

2016-08-24 Thread Uwe Laverenz

Hi Elad,

I sent you a download message.

thank you,
Uwe
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [ovirt-users] iSCSI Multipathing -> host inactive

2016-08-24 Thread Elad Ben Aharon
Network configuration seems OK.
Please provide engine.log and vdsm.log

Thanks

On Wed, Aug 24, 2016 at 3:22 PM, Uwe Laverenz  wrote:

> Hi,
>
> sorry for the delay, I reinstalled everything, configured the networks,
> attached the iSCSI storage with 2 interfaces and finally created the
> iSCSI-bond:
>
> [root@ovh01 ~]# route
>> Kernel IP Routentabelle
>> ZielRouter  Genmask Flags Metric RefUse
>> Iface
>> default hp5406-1-srv.mo 0.0.0.0 UG0  00
>> ovirtmgmt
>> 10.0.24.0   0.0.0.0 255.255.255.0   U 0  00
>> ovirtmgmt
>> 10.0.131.0  0.0.0.0 255.255.255.0   U 0  00
>> enp9s0f0
>> 10.0.132.0  0.0.0.0 255.255.255.0   U 0  00
>> enp9s0f1
>> link-local  0.0.0.0 255.255.0.0 U 1005   00
>> enp9s0f0
>> link-local  0.0.0.0 255.255.0.0 U 1006   00
>> enp9s0f1
>> link-local  0.0.0.0 255.255.0.0 U 1008   00
>> ovirtmgmt
>> link-local  0.0.0.0 255.255.0.0 U 1015   00
>> bond0
>> link-local  0.0.0.0 255.255.0.0 U 1017   00
>> ADMIN
>> link-local  0.0.0.0 255.255.0.0 U 1021   00
>> SRV
>>
>
> and:
>
> [root@ovh01 ~]# ip a
>> 1: lo:  mtu 65536 qdisc noqueue state UNKNOWN
>> link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
>> inet 127.0.0.1/8 scope host lo
>>valid_lft forever preferred_lft forever
>> inet6 ::1/128 scope host
>>valid_lft forever preferred_lft forever
>> 2: enp13s0:  mtu 1500 qdisc pfifo_fast
>> master ovirtmgmt state UP qlen 1000
>> link/ether e0:3f:49:6d:68:c4 brd ff:ff:ff:ff:ff:ff
>> 3: enp8s0f0:  mtu 1500 qdisc mq
>> master bond0 state UP qlen 1000
>> link/ether 90:e2:ba:11:21:d0 brd ff:ff:ff:ff:ff:ff
>> 4: enp8s0f1:  mtu 1500 qdisc mq
>> master bond0 state UP qlen 1000
>> link/ether 90:e2:ba:11:21:d0 brd ff:ff:ff:ff:ff:ff
>> 5: enp9s0f0:  mtu 1500 qdisc mq state
>> UP qlen 1000
>> link/ether 90:e2:ba:11:21:d4 brd ff:ff:ff:ff:ff:ff
>> inet 10.0.131.181/24 brd 10.0.131.255 scope global enp9s0f0
>>valid_lft forever preferred_lft forever
>> 6: enp9s0f1:  mtu 1500 qdisc mq state
>> UP qlen 1000
>> link/ether 90:e2:ba:11:21:d5 brd ff:ff:ff:ff:ff:ff
>> inet 10.0.132.181/24 brd 10.0.132.255 scope global enp9s0f1
>>valid_lft forever preferred_lft forever
>> 7: ;vdsmdummy;:  mtu 1500 qdisc noop state DOWN
>> link/ether 26:b2:4e:5e:f0:60 brd ff:ff:ff:ff:ff:ff
>> 8: ovirtmgmt:  mtu 1500 qdisc noqueue
>> state UP
>> link/ether e0:3f:49:6d:68:c4 brd ff:ff:ff:ff:ff:ff
>> inet 10.0.24.181/24 brd 10.0.24.255 scope global ovirtmgmt
>>valid_lft forever preferred_lft forever
>> 14: vnet0:  mtu 1500 qdisc pfifo_fast
>> master ovirtmgmt state UNKNOWN qlen 500
>> link/ether fe:16:3e:79:25:86 brd ff:ff:ff:ff:ff:ff
>> inet6 fe80::fc16:3eff:fe79:2586/64 scope link
>>valid_lft forever preferred_lft forever
>> 15: bond0:  mtu 1500 qdisc
>> noqueue state UP
>> link/ether 90:e2:ba:11:21:d0 brd ff:ff:ff:ff:ff:ff
>> 16: bond0.32@bond0:  mtu 1500 qdisc
>> noqueue master ADMIN state UP
>> link/ether 90:e2:ba:11:21:d0 brd ff:ff:ff:ff:ff:ff
>> 17: ADMIN:  mtu 1500 qdisc noqueue
>> state UP
>> link/ether 90:e2:ba:11:21:d0 brd ff:ff:ff:ff:ff:ff
>> 20: bond0.24@bond0:  mtu 1500 qdisc
>> noqueue master SRV state UP
>> link/ether 90:e2:ba:11:21:d0 brd ff:ff:ff:ff:ff:ff
>> 21: SRV:  mtu 1500 qdisc noqueue state
>> UP
>> link/ether 90:e2:ba:11:21:d0 brd ff:ff:ff:ff:ff:ff
>>
>
> The host keeps toggling all storage domains on and off as soon as there is
> an iSCSI bond configured.
>
> Thank you for your patience.
>
> cu,
> Uwe
>
>
> Am 18.08.2016 um 11:10 schrieb Elad Ben Aharon:
>
>> I don't think it's necessary.
>> Please provide the host's routing table and interfaces list ('ip a' or
>> ifconfing) while it's configured with the bond.
>>
>> Thanks
>>
>> On Tue, Aug 16, 2016 at 4:39 PM, Uwe Laverenz > > wrote:
>>
>> Hi Elad,
>>
>> Am 16.08.2016 um 10:52 schrieb Elad Ben Aharon:
>>
>> Please be sure that ovirtmgmt is not part of the iSCSI bond.
>>
>>
>> Yes, I made sure it is not part of the bond.
>>
>> It does seem to have a conflict between default and enp9s0f0/
>> enp9s0f1.
>> Try to put the host in maintenance and then delete the iscsi
>> 

Re: [ovirt-users] iSCSI Multipathing -> host inactive

2016-08-24 Thread Uwe Laverenz

Hi,

sorry for the delay, I reinstalled everything, configured the networks, 
attached the iSCSI storage with 2 interfaces and finally created the 
iSCSI-bond:



[root@ovh01 ~]# route
Kernel IP Routentabelle
ZielRouter  Genmask Flags Metric RefUse Iface
default hp5406-1-srv.mo 0.0.0.0 UG0  00 
ovirtmgmt
10.0.24.0   0.0.0.0 255.255.255.0   U 0  00 
ovirtmgmt
10.0.131.0  0.0.0.0 255.255.255.0   U 0  00 enp9s0f0
10.0.132.0  0.0.0.0 255.255.255.0   U 0  00 enp9s0f1
link-local  0.0.0.0 255.255.0.0 U 1005   00 enp9s0f0
link-local  0.0.0.0 255.255.0.0 U 1006   00 enp9s0f1
link-local  0.0.0.0 255.255.0.0 U 1008   00 
ovirtmgmt
link-local  0.0.0.0 255.255.0.0 U 1015   00 bond0
link-local  0.0.0.0 255.255.0.0 U 1017   00 ADMIN
link-local  0.0.0.0 255.255.0.0 U 1021   00 SRV


and:


[root@ovh01 ~]# ip a
1: lo:  mtu 65536 qdisc noqueue state UNKNOWN
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
   valid_lft forever preferred_lft forever
inet6 ::1/128 scope host
   valid_lft forever preferred_lft forever
2: enp13s0:  mtu 1500 qdisc pfifo_fast master 
ovirtmgmt state UP qlen 1000
link/ether e0:3f:49:6d:68:c4 brd ff:ff:ff:ff:ff:ff
3: enp8s0f0:  mtu 1500 qdisc mq master 
bond0 state UP qlen 1000
link/ether 90:e2:ba:11:21:d0 brd ff:ff:ff:ff:ff:ff
4: enp8s0f1:  mtu 1500 qdisc mq master 
bond0 state UP qlen 1000
link/ether 90:e2:ba:11:21:d0 brd ff:ff:ff:ff:ff:ff
5: enp9s0f0:  mtu 1500 qdisc mq state UP qlen 
1000
link/ether 90:e2:ba:11:21:d4 brd ff:ff:ff:ff:ff:ff
inet 10.0.131.181/24 brd 10.0.131.255 scope global enp9s0f0
   valid_lft forever preferred_lft forever
6: enp9s0f1:  mtu 1500 qdisc mq state UP qlen 
1000
link/ether 90:e2:ba:11:21:d5 brd ff:ff:ff:ff:ff:ff
inet 10.0.132.181/24 brd 10.0.132.255 scope global enp9s0f1
   valid_lft forever preferred_lft forever
7: ;vdsmdummy;:  mtu 1500 qdisc noop state DOWN
link/ether 26:b2:4e:5e:f0:60 brd ff:ff:ff:ff:ff:ff
8: ovirtmgmt:  mtu 1500 qdisc noqueue state UP
link/ether e0:3f:49:6d:68:c4 brd ff:ff:ff:ff:ff:ff
inet 10.0.24.181/24 brd 10.0.24.255 scope global ovirtmgmt
   valid_lft forever preferred_lft forever
14: vnet0:  mtu 1500 qdisc pfifo_fast master 
ovirtmgmt state UNKNOWN qlen 500
link/ether fe:16:3e:79:25:86 brd ff:ff:ff:ff:ff:ff
inet6 fe80::fc16:3eff:fe79:2586/64 scope link
   valid_lft forever preferred_lft forever
15: bond0:  mtu 1500 qdisc noqueue 
state UP
link/ether 90:e2:ba:11:21:d0 brd ff:ff:ff:ff:ff:ff
16: bond0.32@bond0:  mtu 1500 qdisc noqueue 
master ADMIN state UP
link/ether 90:e2:ba:11:21:d0 brd ff:ff:ff:ff:ff:ff
17: ADMIN:  mtu 1500 qdisc noqueue state UP
link/ether 90:e2:ba:11:21:d0 brd ff:ff:ff:ff:ff:ff
20: bond0.24@bond0:  mtu 1500 qdisc noqueue 
master SRV state UP
link/ether 90:e2:ba:11:21:d0 brd ff:ff:ff:ff:ff:ff
21: SRV:  mtu 1500 qdisc noqueue state UP
link/ether 90:e2:ba:11:21:d0 brd ff:ff:ff:ff:ff:ff


The host keeps toggling all storage domains on and off as soon as there 
is an iSCSI bond configured.


Thank you for your patience.

cu,
Uwe


Am 18.08.2016 um 11:10 schrieb Elad Ben Aharon:

I don't think it's necessary.
Please provide the host's routing table and interfaces list ('ip a' or
ifconfing) while it's configured with the bond.

Thanks

On Tue, Aug 16, 2016 at 4:39 PM, Uwe Laverenz > wrote:

Hi Elad,

Am 16.08.2016 um 10:52 schrieb Elad Ben Aharon:

Please be sure that ovirtmgmt is not part of the iSCSI bond.


Yes, I made sure it is not part of the bond.

It does seem to have a conflict between default and enp9s0f0/
enp9s0f1.
Try to put the host in maintenance and then delete the iscsi
nodes using
'iscsiadm -m node -o delete'. Then activate the host.


I tried that, I managed to get the iSCSI interface clean, no
"default" anymore. But that didn't solve the problem of the host
becoming "inactive". Not even the NFS domains would come up.

As soon as I remove the iSCSI-bond, the host becomes responsive
again and I can activate all storage domains. Removing the bond also
brings the duplicated "Iface Name" 

Re: [ovirt-users] iSCSI Multipathing -> host inactive

2016-08-18 Thread Elad Ben Aharon
I don't think it's necessary.
Please provide the host's routing table and interfaces list ('ip a' or
ifconfing) while it's configured with the bond.

Thanks

On Tue, Aug 16, 2016 at 4:39 PM, Uwe Laverenz  wrote:

> Hi Elad,
>
> Am 16.08.2016 um 10:52 schrieb Elad Ben Aharon:
>
> Please be sure that ovirtmgmt is not part of the iSCSI bond.
>>
>
> Yes, I made sure it is not part of the bond.
>
> It does seem to have a conflict between default and enp9s0f0/ enp9s0f1.
>> Try to put the host in maintenance and then delete the iscsi nodes using
>> 'iscsiadm -m node -o delete'. Then activate the host.
>>
>
> I tried that, I managed to get the iSCSI interface clean, no "default"
> anymore. But that didn't solve the problem of the host becoming "inactive".
> Not even the NFS domains would come up.
>
> As soon as I remove the iSCSI-bond, the host becomes responsive again and
> I can activate all storage domains. Removing the bond also brings the
> duplicated "Iface Name" back (but this time causes no problems).
>
> ...
>
> I wonder if there is a basic misunderstanding on my side: wouldn't it be
> necessary that all targets are reachable from all interfaces that are
> configured into the bond to make it work?
>
> But this would either mean two interfaces in the same network or routing
> between the iSCSI networks.
>
> Thanks,
>
> Uwe
> ___
> Users mailing list
> Users@ovirt.org
> http://lists.ovirt.org/mailman/listinfo/users
>
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [ovirt-users] iSCSI Multipathing -> host inactive

2016-08-16 Thread Uwe Laverenz

Hi Elad,

Am 16.08.2016 um 10:52 schrieb Elad Ben Aharon:


Please be sure that ovirtmgmt is not part of the iSCSI bond.


Yes, I made sure it is not part of the bond.


It does seem to have a conflict between default and enp9s0f0/ enp9s0f1.
Try to put the host in maintenance and then delete the iscsi nodes using
'iscsiadm -m node -o delete'. Then activate the host.


I tried that, I managed to get the iSCSI interface clean, no "default" 
anymore. But that didn't solve the problem of the host becoming 
"inactive". Not even the NFS domains would come up.


As soon as I remove the iSCSI-bond, the host becomes responsive again 
and I can activate all storage domains. Removing the bond also brings 
the duplicated "Iface Name" back (but this time causes no problems).


...

I wonder if there is a basic misunderstanding on my side: wouldn't it be 
necessary that all targets are reachable from all interfaces that are 
configured into the bond to make it work?


But this would either mean two interfaces in the same network or routing 
between the iSCSI networks.


Thanks,
Uwe
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [ovirt-users] iSCSI Multipathing -> host inactive

2016-08-16 Thread Uwe Laverenz

Hi,

Am 16.08.2016 um 09:26 schrieb Elad Ben Aharon:

Currently, your host is connected through a single initiator, the
'Default' interface (Iface Name: default), to 2 targets: tgta and tgtb


I see what you mean, but the "Iface Name" is somewhat irritating here, 
it does not mean that the wrong interface (ovirtmgmt) is used.
If you have a look at "Iface IPaddress" for both you can see that the 
correct, dedicated interfaces are used:


Iface IPaddress: 10.0.131.122   (iSCSIA network)
Iface IPaddress: 10.0.132.122   (iSCSIB network)


(Target: iqn.2005-10.org.freenas.ctl:tgta and Target:
iqn.2005-10.org.freenas.ctl:tgtb). Hence, each LUN is exposed from the
storage server via 2 paths.
Since the connection to the storage is done via the 'Default' interface
and not via the 2 iSCSI networks you've configured, currently, the iSCSI
bond is not operational.


Please see above. The storage servers iSCSI-addresses aren't even 
reachable from the ovirtmgmt net, they are in completely isolated networks.



For the iSCSI bond to be operational, you'll have to do the following:
- Create 2 networks in RHEVM under the relevant cluster (not sure if
you've already did it) - iSCSI1 and iSCSI2 . Configure both networks to
be non-required networks for the cluster (should be also non-VM networks).
- Attach the networks to the host's 2 interfaces using hosts Setup-networks.
- Create a new iSCSI bond / modify the bond you've created and pick the
2 newly created networks along with all storage targets. Make sure that
the Default network is not part of the bond (usually, the Default
network is the management one - 'ovirtmgmt').
- Put the host in maintenance and re-activate it so the iSCSI sessions
will be refreshed with the new connection specifications.


This is exactly what I did, expect that I had to add the iSCSI-storage 
first, otherwise the "iSCSI Multipathing" tab does not appear in the 
data center section.


I configured an iSCSI-Bond and the problem seems to be that it leads to 
conflicting iSCSI-settings on the host. The host uses the very same 
interface twice only with different "IFace Name":


iSCSIA:

Iface Name: default
Iface Transport: tcp
Iface Initiatorname: iqn.1994-05.com.redhat:cda91b279ac5
Iface IPaddress: 10.0.131.122

Iface Name: enp9s0f0
Iface Transport: tcp
Iface Initiatorname: iqn.1994-05.com.redhat:cda91b279ac5
Iface IPaddress: 10.0.131.122


iSCSIB:

Iface Name: default
Iface Transport: tcp
Iface Initiatorname: iqn.1994-05.com.redhat:cda91b279ac5
Iface IPaddress: 10.0.132.122

Iface Name: enp9s0f1
Iface Transport: tcp
Iface Initiatorname: iqn.1994-05.com.redhat:cda91b279ac5
Iface IPaddress: 10.0.132.122

I guess this is the reason why the host has problems to attach the 
storage domain, it toggles all storage domains on and off all the time.


Thank you,
Uwe
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


Re: [ovirt-users] iSCSI Multipathing -> host inactive

2016-08-16 Thread Elad Ben Aharon
Currently, your host is connected through a single initiator, the 'Default'
interface (Iface Name: default), to 2 targets: tgta and tgtb (Target:
iqn.2005-10.org.freenas.ctl:tgta and Target: iqn.2005-10.org.freenas.ctl:tgtb).
Hence, each LUN is exposed from the storage server via 2 paths.
Since the connection to the storage is done via the 'Default' interface and
not via the 2 iSCSI networks you've configured, currently, the iSCSI bond
is not operational.

For the iSCSI bond to be operational, you'll have to do the following:
- Create 2 networks in RHEVM under the relevant cluster (not sure if you've
already did it) - iSCSI1 and iSCSI2 . Configure both networks to be
non-required networks for the cluster (should be also non-VM networks).
- Attach the networks to the host's 2 interfaces using hosts Setup-networks.
- Create a new iSCSI bond / modify the bond you've created and pick the 2
newly created networks along with all storage targets. Make sure that the
Default network is not part of the bond (usually, the Default network is
the management one - 'ovirtmgmt').
- Put the host in maintenance and re-activate it so the iSCSI sessions will
be refreshed with the new connection specifications.



Please let me know if it works for you.

Elad

On Tue, Aug 16, 2016 at 9:26 AM, Uwe Laverenz  wrote:

> Hi,
>
> Am 15.08.2016 um 16:53 schrieb Elad Ben Aharon:
>
> Is the iSCSI domain that supposed to be connected through the bond the
>> current master domain?
>>
>
> No, it isn't. An NFS share is the master domain.
>
>
> Also, can you please provide the output of 'iscsiadm -m session -P3' ?
>>
>
> Yes, of course (meanwhile I have switched to 2 targets, 1 per portal).
> This is _without_ iSCSI-Bond:
>
> [root@ovh01 ~]# iscsiadm -m session -P3
> iSCSI Transport Class version 2.0-870
> version 6.2.0.873-33.2
> Target: iqn.2005-10.org.freenas.ctl:tgta (non-flash)
> Current Portal: 10.0.131.121:3260,257
> Persistent Portal: 10.0.131.121:3260,257
> **
> Interface:
> **
> Iface Name: default
> Iface Transport: tcp
> Iface Initiatorname: iqn.1994-05.com.redhat:cda91b279ac5
> Iface IPaddress: 10.0.131.122
> Iface HWaddress: 
> Iface Netdev: 
> SID: 34
> iSCSI Connection State: LOGGED IN
> iSCSI Session State: LOGGED_IN
> Internal iscsid Session State: NO CHANGE
> *
> Timeouts:
> *
> Recovery Timeout: 5
> Target Reset Timeout: 30
> LUN Reset Timeout: 30
> Abort Timeout: 15
> *
> CHAP:
> *
> username: 
> password: 
> username_in: 
> password_in: 
> 
> Negotiated iSCSI params:
> 
> HeaderDigest: None
> DataDigest: None
> MaxRecvDataSegmentLength: 262144
> MaxXmitDataSegmentLength: 131072
> FirstBurstLength: 131072
> MaxBurstLength: 16776192
> ImmediateData: Yes
> InitialR2T: Yes
> MaxOutstandingR2T: 1
> 
> Attached SCSI devices:
> 
> Host Number: 44 State: running
> scsi44 Channel 00 Id 0 Lun: 0
> Attached scsi disk sdf  State: running
> scsi44 Channel 00 Id 0 Lun: 1
> Attached scsi disk sdg  State: running
> scsi44 Channel 00 Id 0 Lun: 2
> Attached scsi disk sdh  State: running
> scsi44 Channel 00 Id 0 Lun: 3
> Attached scsi disk sdi  State: running
> Target: iqn.2005-10.org.freenas.ctl:tgtb (non-flash)
> Current Portal: 10.0.132.121:3260,258
> Persistent Portal: 10.0.132.121:3260,258
> **
> Interface:
> **
> Iface Name: default
> Iface Transport: tcp
> Iface Initiatorname: iqn.1994-05.com.redhat:cda91b279ac5
> Iface IPaddress: 10.0.132.122
> Iface HWaddress: 
> Iface Netdev: 
> SID: 35
> iSCSI Connection State: LOGGED IN
> iSCSI Session State: LOGGED_IN
> Internal iscsid Session State: NO CHANGE
> *
> Timeouts:
> *
> Recovery Timeout: 5
> Target Reset 

Re: [ovirt-users] iSCSI Multipathing -> host inactive

2016-08-16 Thread Uwe Laverenz

Hi,

Am 15.08.2016 um 16:53 schrieb Elad Ben Aharon:


Is the iSCSI domain that supposed to be connected through the bond the
current master domain?


No, it isn't. An NFS share is the master domain.



Also, can you please provide the output of 'iscsiadm -m session -P3' ?


Yes, of course (meanwhile I have switched to 2 targets, 1 per portal). 
This is _without_ iSCSI-Bond:


[root@ovh01 ~]# iscsiadm -m session -P3
iSCSI Transport Class version 2.0-870
version 6.2.0.873-33.2
Target: iqn.2005-10.org.freenas.ctl:tgta (non-flash)
Current Portal: 10.0.131.121:3260,257
Persistent Portal: 10.0.131.121:3260,257
**
Interface:
**
Iface Name: default
Iface Transport: tcp
Iface Initiatorname: iqn.1994-05.com.redhat:cda91b279ac5
Iface IPaddress: 10.0.131.122
Iface HWaddress: 
Iface Netdev: 
SID: 34
iSCSI Connection State: LOGGED IN
iSCSI Session State: LOGGED_IN
Internal iscsid Session State: NO CHANGE
*
Timeouts:
*
Recovery Timeout: 5
Target Reset Timeout: 30
LUN Reset Timeout: 30
Abort Timeout: 15
*
CHAP:
*
username: 
password: 
username_in: 
password_in: 

Negotiated iSCSI params:

HeaderDigest: None
DataDigest: None
MaxRecvDataSegmentLength: 262144
MaxXmitDataSegmentLength: 131072
FirstBurstLength: 131072
MaxBurstLength: 16776192
ImmediateData: Yes
InitialR2T: Yes
MaxOutstandingR2T: 1

Attached SCSI devices:

Host Number: 44 State: running
scsi44 Channel 00 Id 0 Lun: 0
Attached scsi disk sdf  State: running
scsi44 Channel 00 Id 0 Lun: 1
Attached scsi disk sdg  State: running
scsi44 Channel 00 Id 0 Lun: 2
Attached scsi disk sdh  State: running
scsi44 Channel 00 Id 0 Lun: 3
Attached scsi disk sdi  State: running
Target: iqn.2005-10.org.freenas.ctl:tgtb (non-flash)
Current Portal: 10.0.132.121:3260,258
Persistent Portal: 10.0.132.121:3260,258
**
Interface:
**
Iface Name: default
Iface Transport: tcp
Iface Initiatorname: iqn.1994-05.com.redhat:cda91b279ac5
Iface IPaddress: 10.0.132.122
Iface HWaddress: 
Iface Netdev: 
SID: 35
iSCSI Connection State: LOGGED IN
iSCSI Session State: LOGGED_IN
Internal iscsid Session State: NO CHANGE
*
Timeouts:
*
Recovery Timeout: 5
Target Reset Timeout: 30
LUN Reset Timeout: 30
Abort Timeout: 15
*
CHAP:
*
username: 
password: 
username_in: 
password_in: 

Negotiated iSCSI params:

HeaderDigest: None
DataDigest: None
MaxRecvDataSegmentLength: 262144
MaxXmitDataSegmentLength: 131072
FirstBurstLength: 131072
MaxBurstLength: 16776192
ImmediateData: Yes
InitialR2T: Yes
MaxOutstandingR2T: 1

Attached SCSI devices:

Host Number: 45 State: running
scsi45 Channel 00 Id 0 Lun: 0
Attached scsi disk sdj  State: running
scsi45 Channel 00 Id 0 Lun: 1
Attached scsi disk sdk  State: running
scsi45 Channel 00 Id 0 Lun: 2
Attached scsi disk sdl  State: running
scsi45 Channel 00 Id 0 Lun: 3
Attached scsi disk sdm  State: running

And `multipath -ll`:

[root@ovh01 ~]# multipath -ll
36589cfc00fafcc87da5ddd69c7e2 dm-2 FreeNAS ,iSCSI Disk

Re: [ovirt-users] iSCSI Multipathing -> host inactive

2016-08-15 Thread Elad Ben Aharon
Hi,

Is the iSCSI domain that supposed to be connected through the bond the
current master domain?

Also, can you please provide the output of 'iscsiadm -m session -P3' ?


Thanks



On Mon, Aug 15, 2016 at 4:31 PM, Uwe Laverenz  wrote:

> Hi all,
>
> I'd like to test iSCSI multipathing with OVirt 4.02 and see the following
> problem: if I try to add an iSCSI-Bond the host loses connection to _all_
> storage domains.
>
> I guess I'm doing something wrong. :)
>
> I have built a small test environment for this:
>
> The storage is provided by a freenas box with two dedicated interfaces for
> two separate iSCSI networks.
>
> Each interface has one address in one network (no VLANs, no trunking). For
> each network there is one portal configured. Both portals point to the same
> target. The target has 4 LUNs.
>
> The host also has two dedicated interfaces for iSCSI and finds the target
> over both portals, all LUNs are selected for use and show 2 paths.
>
> My questions:
>
> 1) Is this setup ok or did I miss something?
>
> 2) The LUNs already show 2 paths (multipath -ll), one "active" and one
> "enabled", what difference would a Datacenter iSCSI-bond make?
>
> 3) What combination of checkboxes do I have to use?
>
> Logical Letworks
>
> [ ] ISCSIA
> [ ] ISCSIB
>
> Storage Targets
>
> [ ] iqn.2005-10.org.freenas.ctl:tgt01   10.0.131.121   3260
> [ ] iqn.2005-10.org.freenas.ctl:tgt01   10.0.132.121   3260
>
>
> As stated in the beginning: all my tests made the host lose connection to
> all storage domains (NFS included) and I can not see what I am doing wrong.
>
> Thank you very much!
>
> cu,
> Uwe
> ___
> Users mailing list
> Users@ovirt.org
> http://lists.ovirt.org/mailman/listinfo/users
>
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users


[ovirt-users] iSCSI Multipathing -> host inactive

2016-08-15 Thread Uwe Laverenz

Hi all,

I'd like to test iSCSI multipathing with OVirt 4.02 and see the 
following problem: if I try to add an iSCSI-Bond the host loses 
connection to _all_ storage domains.


I guess I'm doing something wrong. :)

I have built a small test environment for this:

The storage is provided by a freenas box with two dedicated interfaces 
for two separate iSCSI networks.


Each interface has one address in one network (no VLANs, no trunking). 
For each network there is one portal configured. Both portals point to 
the same target. The target has 4 LUNs.


The host also has two dedicated interfaces for iSCSI and finds the 
target over both portals, all LUNs are selected for use and show 2 paths.


My questions:

1) Is this setup ok or did I miss something?

2) The LUNs already show 2 paths (multipath -ll), one "active" and one 
"enabled", what difference would a Datacenter iSCSI-bond make?


3) What combination of checkboxes do I have to use?

Logical Letworks

[ ] ISCSIA
[ ] ISCSIB

Storage Targets

[ ] iqn.2005-10.org.freenas.ctl:tgt01   10.0.131.121   3260
[ ] iqn.2005-10.org.freenas.ctl:tgt01   10.0.132.121   3260


As stated in the beginning: all my tests made the host lose connection 
to all storage domains (NFS included) and I can not see what I am doing 
wrong.


Thank you very much!

cu,
Uwe
___
Users mailing list
Users@ovirt.org
http://lists.ovirt.org/mailman/listinfo/users