Tommy Downtime

[flazepe]

Yes. I don't see a reason for it to not work.

[jamieoz]

G'day

 

Just to let you know, I can live without my actual websites until Tommy is back up, but I would like to be able to access my e-mails at the 2 domains that I have pointing to Tommy. 

 

Is this possible and how do I go about doing it?

 

Regards

 

Jamie.

[Luigi123]

No, you cannot access your emails while Tommy is down, unfortunately.

[wolstech]

Mail is in your backup. There should be a mail folder in your home folder backup with your messages. They won’t have meaningful names though and are “raw” (they have the headers attached) so you’d need to sort them.

[smac992]

As far as "transaction ID" mentioned here:

 

3. If you're a donor, you can request a Ricky invite. Please create a new topic in customer service with the transaction ID of your most recent donation (it can be the same one you originally used for Tommy, even if it was long ago), and your Tommy username.

 

Would that be the validation code in the URL from the link in the "thank you for sending us money..." email that I got when I signed up for Tommy back in 2016, or something else?  

 

Thanks.

[Luigi123]

No, we only need your recent transaction ID. If you donate it using GoFundMe before, provide your email address. Then, once our Root Admin verifies that, they will send you an invitation for Ricky Server to your email.

[Wooldrid]

Morning, I am battling to create a temp account until Tommy is rebuilt, I've got my backup (thanks) but my timezone is such that The servers new Account cap is reached by 04:00 in the morning. Is their a way to cue a new account request.  On a separate note will Tommy's IP remain the same

[Sn1F3rt]

You can make a donation (minimum $1) and get a Ricky invite.

 

Yeah, Tommy's IP will be remaining the same.

[DmC]

Tommy is alive:

 

https://tommy.heliohost.org:2083/

 

I just got a notification that certbot renew certificates for my domains

 

Credentials don't work though

 

I guess we're getting there.

Edited August 23, 2019 by DmC

[Sn1F3rt]

Oh yeah DmC! Working for me too.

 

Your previous credentials won't work bcoz Tommy's been formatted.

[flexo]

 

Those data are registered internally on the disk and are readable through SMART.

The server runs ESXi free edition, which does not provide such capability (except showing a table by using a rather long sequence of shell commands). All of the storage API functions that would allow programs inside the VMs to easily pull this information are paid features. ESX is not cheap if you use any of the paid versions...$600/year for the cheapest license (vSphere Essentials I think it's called).

 

 

A virtual machine guest that can get information about its host hardware is a bad idea in general...

 

For most enterprise storage controllers, there is a vendor-supplied CIM provider for monitoring hardware health in the ESXi vSphere client interface. This can either be installing the VIBs for hardware support (populates the "Health Status" part of the vSphere management), using and maintaining vendor customized image and patches for such support, or third-party tools such as Dell OpenManage Server for ESXI.

 

. There is also out-of-channel management, such as Dell iDRAC, that can provide storage array health and configuration. This can be monitored using SNMP. Some of these are license upgrades on your hardware, though.

 

Also vendor tools for VMWare like perccli for Dell PERC RAID controllers to get you some text such as online status, and can be used to issue consistency checks.

 

For directly-connected SATA, I just typed in a short sequence of shell commands to my unmanaged disk ESXi:

[root@vmware:~] esxcli storage core device list | grep '  Display Name:' | cut -d'(' -f2 | cut -d')' -f1 | while read DISK
> do
>  echo ": $DISK :"
>  esxcli storage core device smart get -d $DISK
> done
: t10.ATA_____WDC_WD2502ABYS2D02B7A0________________________WD2DWCAT1C903743 :
Parameter                     Value  Threshold  Worst
----------------------------  -----  ---------  -----
Health Status                 OK     N/A        N/A
Media Wearout Indicator       N/A    N/A        N/A
Write Error Count             N/A    N/A        N/A
Read Error Count              0      51         N/A
Power-on Hours                16     0          16
Power Cycle Count             65     0          N/A
Reallocated Sector Count      0      140        N/A
Raw Read Error Rate           0      51         N/A
Drive Temperature             98     0          80
Driver Rated Max Temperature  N/A    N/A        N/A
Write Sectors TOT Count       0      0          N/A
Read Sectors TOT Count        0      0          N/A
Initial Bad Block Count       N/A    N/A        N/A
: t10.ATA_____WDC_WD40EZRX2D00SPEB0_________________________WD2DWCC4E5ZK0N4C :
Parameter                     Value  Threshold  Worst
----------------------------  -----  ---------  -----
Health Status                 OK     N/A        N/A
Media Wearout Indicator       N/A    N/A        N/A
Write Error Count             N/A    N/A        N/A
Read Error Count              0      51         N/A
Power-on Hours                68     0          68
Power Cycle Count             67     0          N/A
Reallocated Sector Count      0      140        N/A
Raw Read Error Rate           0      51         N/A
Drive Temperature             44     0          N/A
Driver Rated Max Temperature  N/A    N/A        N/A
Write Sectors TOT Count       0      0          N/A
Read Sectors TOT Count        0      0          N/A
Initial Bad Block Count       N/A    N/A        N/A

I suppose it is up to the more clever script programmer to figure out how to use such SMART data, such as watching for increasing reallocated sectors. When a SATA hard drive is going to fail, making guesses based on SMART is not quite as good as RAID 10 with a hot spare...or ZFS on your passthru storage drives with multiple drive software RAID redundancy, snapshots etc.

[wolstech]

I've used OpenManage Server Admin in the past on ESXi. It's been a while though since I've last had to set it up. Does the job quite nicely when used alongside the Dell ESXi image. OMSA would work on Sparkie if Krydos wanted to set it up since he's a PowerEdge, but Eddie is HP hardware (DL360 G6).

 

The console commands for the tables in your last example are the only thing I've used recently.

 

Proper redundancy is nice (and when I do servers, I do exactly what you said when I can...RAID 10 with a spare), but historically we've preferred capacity over reliability for cost and ad revenue reasons. It was only recently when we got the NAS that we got our first hardware with a proper RAID (I forget what type, I didn't set it up). Sparkie I believe was supposed to be configured as a raid as well (I think 10 was the preference, but we may have gone with 5). 

 

More often than not though, when our servers fail it's not hardware anyway, it's software corruption. cPanel was never designed to handle what we do to it. It's rated for a few hundred accounts (I think they say 400 max) per server, with the occasional create or delete. We cram well over a thousand in there (Tommy had around 2k when he died of disk failure, Johnny currently has 1500 active accounts, 10k if you count archives), and we're adding and removing users daily. In fact, cP themselves has admitted to us that we're the most extreme use case they have. Their support is always both dumbfounded and impressed that we've managed to make it work whenever we contact them (granted we do modify some parts of cP quite a bit to make it work, most of it is stock), then surprised when we tell them we give the accounts away for free on top of it all (they've given us tons of discounts for this...we're basically free marketing for cP). If anything, credit here goes to cPanel for a product that's well built to the point where can easily exceed its rated specs, and Krydos for spending all the time making it work!

[flexo]

HP RAID controller talk:

https://www.reddit.com/r/homelab/comments/6pppd0/hp_p410i_raid_on_esxi_65_host_any_way_to_view_data/

 

It's usually pretty safe to incrementally apply HP patches on top of their custom ESXi images. Manually installing drivers has a higher chance of bork. I'm guessing the window to mess with ESXi in-person was just missed, and the benefits of health monitoring would be limited in JBOD anyway. As I posted elsewhere, one wonders if the drive will turn out to be good, could have simply been spun back up, and just dropped out from load-based TLER timeout that hasn't been addressed.

 

More web server VMs on the same hardware and drives might be a solution to some concerns, if that doesn't double the per-user cPanel costs (like published metal pricing does), and you've got memory. ESXi can additionally enforce resource limits, so less users are impacted by a bad actor. Automating and obfuscating the server assignment away from users, as most hosting providers do, may be the missing component.

[neci]

Thank you heliohost for Tommy, good his back online.

 

Now please how can I login? As it seems my password is not working? Was it reset, probably yes due to the server reinstalling.

 

Is there a new default password? Or a link to login to cpanel? I won't to go on uploading the website again. Thanks

[flazepe]

Tommy is not ready for sign ups yet. You'll need to wait until we make an announcement.