Installation issue...

[viral anchor]

Setting up a new Grid with 6 storage nodes 5728... They were all added like normal and we ended up at the last part of the installation... and they now have a status of Preparing Storage... and if I look at the nodes installation page is states "The Installation Hasn't Been Started"... but I has and I did... 🙂 My question is, if I reboot the nodes... will they try to continue their installation? I can also "Home" and start a new installation... but I am not sure if they will be added into the list... and is it even possible to "remove" the two nodes from the installation at this part and join them later? Any ideas? Or is it a whole re-install at this point?

...This is SG version 12.0 btw 🙂

Here is what is shown on one of the last nodes...

[viral anchor]

...tried a reboot of one of the "hanging" storage nodes... just booted up to the same state... if I look on the Santricity page it shows that the pool and LUNs have been created correctly...

[viral anchor]

OK I managed to get one of the two storage nodes further to where it waits for Cassandra like the others... This was by starting the installation like we already did... it skipped the first parts, but downloaded the OS and installed OK... Yet the last node didn't fair so good... it did the same installation, then rebooted to "bare metel"... but then nothing... no web either.. is there some way to reach it, or do I have to be on site to re-install it?

...never seen this process this "flakey" before... had I know I would have started with just the minimum 3 nodes...

[viral anchor]

hmmm the last node keeps getting back to the Applicance Installer... it does all the installation etc.. and boots into "bare metal" mode, but eventually ends up again in the installer... just great 🙂 any idea how to see any error logs etc. ?

[viral anchor]

...OK I think it has created too many LUNs in the e-series pool because I have a warning on that... so I have now tried to clear the e-series and trying once again to install the node....

[viral anchor]

OK I have now tried several times to clear the e-series array setup which boots the storage-ctl... I then clear the array, so that it is 100% OK with no errors and no pool etc.. Then I reboot the storage-grid node... verify all network and IPs... then start the installation... the Admin Install actually picks up on the attempt and lists the progress... it all ends with "booting into bare metal..." this takes a looong time... and it eventually ends up at "Preparing storage"... and the Appliance install is avaliable again... no errors, no nothing... no apparent description of how to access the node via ssh when in the installer GUI... so I'm kinda stuck... I think I will need to get onsite, and nuke all 6 nodes again.. then install the SG with just 3 storage nodes... and try to see if I can add the others one at a time...

[gusty river]

yeah, there's an ssh login to the installer image. You can get the credentials from support.

[viral anchor]

I finally got it working... after 10-15 attempts... I managed to login to the node with SSH, but there isn't much to see there... it was like it installed everything, then rebooted into baremetal mode... then after 15-30 min. it reuturned again into the installed/maint mode... The last thing I did was to set the e-series IP address to DHCP (the cables are not connected). Then cleared the e-series, rebooted the e-series, rebooted the storage node.. then did the install again.. and it finally worked... I have done all of the above before except the DHCP, which seemed strange that it could have any effect? Anyway I now have a glorious new Grid 🙂 ....and first thing I see this morning is that one disk is failing in the "problem" node.. so maybe this could also be the cause?... It would be nice if they could allow you to remove nodes as part of the installation... just to get the Grid going... I feared the worst where I had to clear all nodes and try again... which would involve a site visit etc..