Discussion:
[Linux-ha-dev] announcement: resource-agents release candidate 3.9.5rc1
Dejan Muhamedagic
2013-01-30 10:30:39 UTC
Permalink
Hello,

The current resource-agents repository has been tagged
v3.9.5rc1. It is mainly a bug fix release.

The full list of changes for the linux-ha RA set is available in
ChangeLog:

https://github.com/ClusterLabs/resource-agents/blob/v3.9.5rc1/ChangeLog

We'll allow a week for testing. The final release is planned for
Feb 6.

Many thanks to all contributors!

Best,

The resource-agents maintainers
Keisuke MORI
2013-01-31 09:33:49 UTC
Permalink
Hi,

Does IPaddr2 need to support 'eth0:label' format in a single 'nic'
parameter when you want to use an interface label?

I thought it does't from the mata-data description of 'nic' but it
looks conflicting with the 'iflabel' description;

nic:
Do NOT specify an alias interface in the form eth0:1 or anything here;
rather, specify the base interface only.
If you want a label, see the iflabel parameter.

iflabel:
(omit)
If a label is specified in nic name, this parameter has no effect.


The latest IPaddr2 (findif.sh version) would reject it as an invalid
nic has specified.
If we do need to support it I will submit a patch for this.

Thanks,
Post by Dejan Muhamedagic
Hello,
The current resource-agents repository has been tagged
v3.9.5rc1. It is mainly a bug fix release.
The full list of changes for the linux-ha RA set is available in
https://github.com/ClusterLabs/resource-agents/blob/v3.9.5rc1/ChangeLog
We'll allow a week for testing. The final release is planned for
Feb 6.
Many thanks to all contributors!
Best,
The resource-agents maintainers
_______________________________________________
ha-wg-technical mailing list
https://lists.linuxfoundation.org/mailman/listinfo/ha-wg-technical
--
Keisuke MORI
Dejan Muhamedagic
2013-01-31 09:53:53 UTC
Permalink
Hi Keisuke-san,
Post by Keisuke MORI
Hi,
Does IPaddr2 need to support 'eth0:label' format in a single 'nic'
parameter when you want to use an interface label?
I thought it does't from the mata-data description of 'nic' but it
looks conflicting with the 'iflabel' description;
Do NOT specify an alias interface in the form eth0:1 or anything here;
rather, specify the base interface only.
If you want a label, see the iflabel parameter.
(omit)
If a label is specified in nic name, this parameter has no effect.
Hmpf.
Post by Keisuke MORI
The latest IPaddr2 (findif.sh version) would reject it as an invalid
nic has specified.
If we do need to support it I will submit a patch for this.
I'd rather just update the iflabel description. After all,
normally one doesn't need to specify the nic. But you'll
get different preferences from different people :)
However, it seems to be a regression, so we should probably allow
labels. BTW, is this related to
https://github.com/ClusterLabs/resource-agents/issues/200 ?

Cheers,

Dejan
Post by Keisuke MORI
Thanks,
Post by Dejan Muhamedagic
Hello,
The current resource-agents repository has been tagged
v3.9.5rc1. It is mainly a bug fix release.
The full list of changes for the linux-ha RA set is available in
https://github.com/ClusterLabs/resource-agents/blob/v3.9.5rc1/ChangeLog
We'll allow a week for testing. The final release is planned for
Feb 6.
Many thanks to all contributors!
Best,
The resource-agents maintainers
_______________________________________________
ha-wg-technical mailing list
https://lists.linuxfoundation.org/mailman/listinfo/ha-wg-technical
--
Keisuke MORI
_______________________________________________________
http://lists.linux-ha.org/mailman/listinfo/linux-ha-dev
Home Page: http://linux-ha.org/
Keisuke MORI
2013-01-31 10:43:56 UTC
Permalink
Hi Dejan,
Post by Dejan Muhamedagic
Post by Keisuke MORI
The latest IPaddr2 (findif.sh version) would reject it as an invalid
nic has specified.
If we do need to support it I will submit a patch for this.
I'd rather just update the iflabel description.
Me too, but
Post by Dejan Muhamedagic
After all,
normally one doesn't need to specify the nic. But you'll
get different preferences from different people :)
However, it seems to be a regression, so we should probably allow
labels.
Yes, that is true. I will submit a patch tomorrow.
Post by Dejan Muhamedagic
BTW, is this related to
https://github.com/ClusterLabs/resource-agents/issues/200 ?
Yes, the proposed patch in #200 would not support the nic:label format
either so it should be re-written in a different way. Honestly I'm
wondering if it is really worth to fix #200 though.

Thanks,
--
Keisuke MORI
Loading...