Real-time performance monitoring and troubleshooting
Post Reply
OREOSpeedwagon
Influencer
Posts: 12
Liked: never
Joined: Jul 29, 2010 1:48 am
Full Name: OREOSpeedwagon
Contact:

Observations/Wish List for 5.0

Post by OREOSpeedwagon »

First off, thanks to the Veeam team to being so quick to follow up on my questions! I have another thread started but the topic started to drift from the title, so I thought I'd start a new one with my observations about 5.0 and my wish list for it.

Here's a few things I notice:

Browsing datastores under the "datastore" view - It would be nice to see all of the VMs in a datastore from this view. Now, I can see the hosts connected to it and get some general metrics on read/write latency, but from this view I can't see what VMs are in that datastore or then see what their performance metrics look like relative to each other within that datastore. It would be nice to have under the "Disk I/O" section (or a separate section) a listing of the VMs in that datastore where you could then see various performance metrics related to storage.

Browsing datastores under the individual host view - On the "Disk" tab it lists the disks by their naa. name. How about changing that display to the volume's datastore name, exactly how it is under the "datastore" view. Or, at least have the option. That way I don't have to jot down the naa to datastore name mapping while I'm troubleshooting or reviewing the data.

Proactive health monitoring - OK, now I'm being greedy... :D But, I like the idea of an automated health check as opposed to just up/down alarms and monitoring. It would be nice if the tool would go through and test the system like some of the scripts out there do, sort of like an automated version of this - http://sourceforge.net/projects/esxhealthscript/

BTW, I'm not being harsh or trying to find fault. I like this tool! I've been looking at this one along w/ Hyper9 and vFoglight and a few others. Each has their sweet spot, but the things I'm pointing out are the sorts of things that I would use on a daily basis and would find to be material differentiators between products.

Let me know when it's all done and I'll buy a copy.

;-)
Vitaliy S.
VP, Product Management
Posts: 27055
Liked: 2710 times
Joined: Mar 30, 2009 9:13 am
Full Name: Vitaliy Safarov
Contact:

Re: Observations/Wish List for 5.0

Post by Vitaliy S. »

Actually, if you try to stand on the particular datastore and choose Stack by option under the graphs and switch from ESX(i) Server to Virtual Machine view, you should see all the VMs running on the datastore and get all Disk I/O statistics for each VM.

In addition, as a workaround you may trigger the script mentioned above on any ESX(i) Server event, by creating your custom alarm or editing the existing alarms. Additionally, all the issues happened on the ESX(i) Server could be quickly observed on the Summary page for particular ESX(i) Server. But, again thanks for the feedback!
OREOSpeedwagon
Influencer
Posts: 12
Liked: never
Joined: Jul 29, 2010 1:48 am
Full Name: OREOSpeedwagon
Contact:

Re: Observations/Wish List for 5.0

Post by OREOSpeedwagon »

Vitaliy S. wrote:Actually, if you try to stand on the particular datastore and choose Stack by option under the graphs and switch from ESX(i) Server to Virtual Machine view, you should see all the VMs running on the datastore and get all Disk I/O statistics for each VM.

In addition, as a workaround you may trigger the script mentioned above on any ESX(i) Server event, by creating your custom alarm or editing the existing alarms. Additionally, all the issues happened on the ESX(i) Server could be quickly observed on the Summary page for particular ESX(i) Server. But, again thanks for the feedback!
Hmm... I'm looking at the GUI in the 5.0 beta and I don't see the "stack by" option.

Also, here's another "want"... Display all VMs in a list view w/ the "greatest hits" of performance metrics. It would be nice to just click on my datacenter and then show a list of all VMs (like I can in vCenter). From there, it would be nice to see a simple list view of: All VMs, amount of RAM/Disk/Network being consumed, snapshot size if existent, and maybe even the last alarm. I think this is all data collected by the tool, it's just a question of surfacing it.

Thoughts?
OREOSpeedwagon
Influencer
Posts: 12
Liked: never
Joined: Jul 29, 2010 1:48 am
Full Name: OREOSpeedwagon
Contact:

Re: Observations/Wish List for 5.0

Post by OREOSpeedwagon »

One more thing, and I may have already mentioned this somewhere else... When looking at the Datastores view in 5.0, if I look under "Disk I/O" I see a listing of all of my hosts. Under "chart options" I have selected my datastores, but the GUI still just displays my hosts. See disk I/O counters, such as latency, for hosts is handy, but what would make more sense is to see the actual VMFS volumes listed w/ the disk I/O counters. From there, click on the VMFS store and then see a listing of the VMs in that store and their I/O counters. This speaks to the notion of better and more focused visability, and tighter drill down/through.
Vitaliy S.
VP, Product Management
Posts: 27055
Liked: 2710 times
Joined: Mar 30, 2009 9:13 am
Full Name: Vitaliy Safarov
Contact:

Re: Observations/Wish List for 5.0

Post by Vitaliy S. »

Hi Pete,

As regards greatest hits of performance metrics for VMs, this information is available right now if you stand on the Top VMs tab. It will dispay you top consumers for CPU, Memory, Disk and Network. Furthermore, if you stand on the Overview tab for the cluster/datacenter node you will be able to see top consumers for CPU and Memory.

Seeing actual VMFS volumes on the Datastores object makes perfect sence for me either. I will see what we can do here. Thanks for a great feedback!
OREOSpeedwagon
Influencer
Posts: 12
Liked: never
Joined: Jul 29, 2010 1:48 am
Full Name: OREOSpeedwagon
Contact:

Re: Observations/Wish List for 5.0

Post by OREOSpeedwagon »

Yeah, the "Top VMs" tab is kind of what I'm looking for. It would be nice if it were all in one column, like the VI client, for easier sorting. Also, it doesn't have any snapshot info on there.
Vitaliy S.
VP, Product Management
Posts: 27055
Liked: 2710 times
Joined: Mar 30, 2009 9:13 am
Full Name: Vitaliy Safarov
Contact:

Re: Observations/Wish List for 5.0

Post by Vitaliy S. »

The feature request for displaying VMFS volumes with disk I/O counters under Datastores view has been implemented and will make it to the RTM code. Thanks again for the feedback!
Vitaliy S.
VP, Product Management
Posts: 27055
Liked: 2710 times
Joined: Mar 30, 2009 9:13 am
Full Name: Vitaliy Safarov
Contact:

Re: Observations/Wish List for 5.0

Post by Vitaliy S. »

OREOSpeedwagon wrote:... Also, it doesn't have any snapshot info on there.
Just a quick update - Top Snapshot users will also be displayed on the Top VMs tab along with Top CPU/Memory/Disk/etc. consumers in the v5. Thanks!
jgremillion
Enthusiast
Posts: 87
Liked: never
Joined: Oct 20, 2009 2:49 pm
Full Name: Joe Gremillion
Contact:

Re: Observations/Wish List for 5.0

Post by jgremillion »

Vitaliy S. wrote: Just a quick update - Top Snapshot users will also be displayed on the Top VMs tab along with Top CPU/Memory/Disk/etc. consumers in the v5. Thanks!
That's a very nice addition! Two great Suggestions! BTW - Vitaly, when will Monitor 5.0 be released?
Vitaliy S.
VP, Product Management
Posts: 27055
Liked: 2710 times
Joined: Mar 30, 2009 9:13 am
Full Name: Vitaliy Safarov
Contact:

Re: Observations/Wish List for 5.0

Post by Vitaliy S. »

Joe, we are planning to release it shortly, so stay tuned!
Post Reply

Who is online

Users browsing this forum: No registered users and 4 guests