Preview, Feedback wanted - Cluster Dashboard

[dcsapak]

Regarding the "what to display"; I'd like to see an information per node on "Current PVE/Software version", as well as an information on if there are patches available or not, if possible distinguished by "regular maintenance" or "urgent security issue" (maybe based on a color etc.). Also I'd like to get some information if there are maybe errors or alerts on a node (little exclamation mark?) that need immediate attention, as well an information on the uptime of each node.

we already have some plans for something like a "notification area" which would display things like fencing events, available updates, etc.
but our todo list is already long ,

current version is not a simple thing, because we would have to make an api call for each node, and i really want to avoid that
(incorporating the version into the /cluster/resources call would be a solution, but i think this is impractical and would bloat it unnecessarily)

uptime should be no problem

My only request is that the information stored on this page can be queried via API. This would make it easy to automate auto-balancing and scaling across an environment, as well as monitoring your cluster at a high level.

i already use the api call, by accumulating the values from the /cluster/resources api call
another api call which basically does the same makes no sense to me really

It would be great if you can incorporate info like ceph-dash (Ceph Cluster Placement Group Status) too.
Just a small column or value is sufficient. Even though this can be monitored from Proxmox GUI, but we relied a lot on ceph-dash to see the PG status, especially when re-balancing in place.

after the cluster dashboard, i am onto reworking the ceph status dashboard of the nodes

 

[Micsi]

Does the dashboard change to show the information of a single node when I click one in the bottom list? That would be sort of a drill down detail view reusing existing views.

Anyway very nice and promising. Looks very helpful!

 

[dcsapak]

Does the dashboard change to show the information of a single node when I click one in the bottom list? That would be sort of a drill down detail view reusing existing views.

No, it does not, but i keep that feature in mind for a future version. (but the only information you get, would be the number of vms/containers running/stopped/etc... i don't know if this is interesting?)

 

[Micsi]

No, it does not, but i keep that feature in mind for a future version. (but the only information you get, would be the number of vms/containers running/stopped/etc... i don't know if this is interesting?)

I would expect visual representation of ressource allocation for the node. RAM, CPU, Storage. That would be interesting.

 

[Mark van de Louw]

Wow that will be an amasing feature! As some others have said, my daily task is hopping around on all hosts to check the usages, this will be a great solution!

Is there an estimate release date?

 

[fireon]

Wow that will be an amasing feature! As some others have said, my daily task is hopping around on all hosts to check the usages, this will be a great solution!
Is there an estimate release date?

Interresting too. Looks like amazing

 

[dcsapak]

fyi: the first version of the dashboard is now on pvetest

 

[morph027]

Going to update my lab box now

 

[Alessandro 123]

Please don't forget about gluster.
A gluster status like for ceph is really usefull

 

[robhost]

Any chance to get network throughput or other netstats in there?

 

[morph027]

Looks good so far....does this implement extra API calls for quering stuff? Just because i'm trying to get the Proxmox cluster into Check_MK (nobody is looking at dashboards all the time )

 

[dcsapak]

Looks good so far....does this implement extra API calls for quering stuff?

no just using
/cluster/resources
/cluster/status
/nodes/localhost/ceph/status

but i calculate the data in the client

 

[morph027]

Idea:

When Ceph isn't in OK state, either directly include some info about it (popup, ...) or link to one of the Ceph management pages which are being displayed at node level...

 

[lilCDNnrg]

Just created a test set of nodes in my virtual lab and I am loving the new dashboard. I can't wait for it to be "production" ready for my live use nodes.

 

[fireon]

Thanks for the first version in pvenosubscriptionrepo. Really nice But two things immediately occurred to me. First thing the storage view. I see it combined all storages, with this i have a problem when i use ZFS. On this i have more then one dataset on the local server. In my case here 4.
- kvm hdd
- kvm ssd
- lxc hdd
- lxc ssd

So i see much more store that i really have.

The second one not so importand but worth mentioning: When you scale the webbrowser to 110 or 125% some datas are cutted. This should be fixed. In this time with monitors with really big resolutions a webpagescaling must be possible.

Thanks a lot

 

[fabian]

Thanks for the first version in pvenosubscriptionrepo. Really nice But two things immediately occurred to me. First thing the storage view. I see it combined all storages, with this i have a problem when i use ZFS. On this i have more then one dataset on the local server. In my case here 4.
- kvm hdd
- kvm ssd
- lxc hdd
- lxc ssd

So i see much more store that i really have.

but this is to be expected if you define multiple storages on the same physical storage - you will always see "more" available space than you can actually use. this is not limited to ZFS, you can do the same with directory, NFS, LVM or combinations of those (and probably all / most of the others as well )

 

[grin]

We throw all kinds of stuff into graphite (using collectd mostly) and visualise it with grafana (so this feature is just a "nice to have" to me), but as a general note: we always prefer to have historical data (graph / rrd) to gauges, usually much more useful than current values. When I display a gauge I usually put the graph in its background.

 

[dcsapak]

i still have the "which storages are counted" feature on my todo list, just not enough time

 

[czechsys]

Cluster status (online/offline nodes) can be optically different than Server View, check image in https://forum.proxmox.com/threads/losing-network-on-one-node-from-3-node-cluster.30696/

 

[dcsapak]

Cluster status (online/offline nodes) can be optically different than Server View, check image in https://forum.proxmox.com/threads/losing-network-on-one-node-from-3-node-cluster.30696/

yes, because for the online/offline count we use the corosync data from /cluster/status and
for the tree we use the resource data from /cluster/resources (which reads /etc/pve/.rrd, which gets updated by the pvestatd)

but we are trying to improve the pvestatd architecture, to better reflect the real status in the future, but this is sadly not so easy