StarWind road map

[robnicholson]

Is there a road map for development of StarWind? I could be dreaming but I'm sure I read something about monitoring. Could have been in these very forums.

Cheers, Rob.

[anton (staff)]

We don't publish official roadmap as we do notiriously change features (represent new and fail old ones) on every day basis

Yes, Monitor is included with V5.7 so you're welcomed to participate in Beta program or keep your hands clean & ready for next couple of weeks if you wish to deal with "gold" version entirely. It works this way: we're represented basic mechanism and GUI and crafted some initial monitoring features. You do tell us what you think you need to have reported thru GUI and alerted - and we add them

Is there a road map for development of StarWind? I could be dreaming but I'm sure I read something about monitoring. Could have been in these very forums.

Cheers, Rob.

[robnicholson]

Might take you up on the beta program except we are rather busy right now...

Two from my wish list:

1. Some form of progress bar when carrying out full-clone
2. Some "in-place" snapshot features like flatten out all snapshots & revert to snapshot

Cheers, Rob.

[Alex (staff)]

Rob,
Thank you for your suggestions!

1. We are working on this functionality now. It will be implemented in 5.7 release.
2. Now it is implemented as a merge for pair of snapshots or delete operation for one snapshot. We will add "flatten all snapshots" and "revert to snapshot" commands in later versions.

[kmax]

One suggestion I have is...

Say Node 2 is complete failure. As in replace the box or redo the physical disk scenario.

It would be nice if you could start Node 2 from scratch...and "join" it with Node 1, and have it start automatically replicating the HA features without the need to recreate targets and copy the img files. Which means you wouldn't need to take Node 1 offline.

[Alex (staff)]

kmax,
we are already working on this feature, it's planned for 5.8 release.
Thank you!

[kmax]

kmax,
we are already working on this feature, it's planned for 5.8 release.
Thank you!

Awesome.

If you combined this, the ability to extend HA target size, snapshotting, and modify cache settings there wouldn't much else to complain about.

[anton (staff)]

You'll always have something to complain about. And it's normal. You point - we fix. You ask - we implement. We call this customer service

kmax,
we are already working on this feature, it's planned for 5.8 release.
Thank you!

Awesome.

If you combined this, the ability to extend HA target size, snapshotting, and modify cache settings there wouldn't much else to complain about.

[anton (staff)]

One question so far... To you and rest of the guys (that's why asked in public).

How do you want HA device to grow? Do you want us to just extend volume capacity (say start reporting 10TB instead of 4TB reported before) in a proper way. Or can we represent another LUN? So
first 4TB would be LUN0 and new 6TB would become as LUN1 on the same HA device? Logical Volume Manager would allow to grow file system and span volume in both cases. So?

[hfourie]

My opinion, muliple luns "bonded" can become confusing. One LUN should just be that, one LUN.

[anton (staff)]

OK, your voice is accepted

My opinion, muliple luns "bonded" can become confusing. One LUN should just be that, one LUN.

[deiruch]

According to the Microsoft-iSCSI-Logo-Requirements-1.0.doc:

"If the size of a logical unit is changed, the target must report a Unit Attention condition of (06/3F/0A) VOLUME SET CREATED OR MODIFIED or (06/28/00) NOT READY TO READY CHANGE, MEDIUM MAY HAVE CHANGED"

I never tested it but it sounds like the proper way to do this. That way you could resize the LUN on the fly. Wouldn't be suprised however if MS Clustering didn't like that (or vmWare, ...).

[anton (staff)]

Correct. Another way is still report different number of LUNs to REPORT_LUNS command sent.

MS would work this way and VMware would not.

According to the Microsoft-iSCSI-Logo-Requirements-1.0.doc:

"If the size of a logical unit is changed, the target must report a Unit Attention condition of (06/3F/0A) VOLUME SET CREATED OR MODIFIED or (06/28/00) NOT READY TO READY CHANGE, MEDIUM MAY HAVE CHANGED"

I never tested it but it sounds like the proper way to do this. That way you could resize the LUN on the fly. Wouldn't be suprised however if MS Clustering didn't like that (or vmWare, ...).

[hfourie]

Hi,

How can I go about testing the Beta for your new version?

[anton (staff)]

Apply for beta program and drop a message to support@starwindsoftware.com telling you would like to have pre-release stuff (in case beta team would be late with e-mail blast).

However we're close to V5.7 release and V5.8 is not ready for beta stage yet. So you'll have 2-3-4 weeks of delay in any case )

Hi,

How can I go about testing the Beta for your new version?