Posts made by Net Runner

@IRJ said in New VM keeps turning off:

You might consider setting up some type of monitoring for it temporarily

I've had a similar issue a while ago that was caused by a memory leak in Exchange itself. It kept consuming more and more RAM and at the end of the day crashed the whole VM with a blue screen. It was happening like once a week usually on weekends thus I didn't pay too much attention to the issue since nobody complained. However, it got worth with time so I had to investigate what is actually going on and take measures like adding more ram, defragging the DB and so on.

Drop some monitoring agent inside the Exchange VM like Veeam One Monitor to be able to understand what's going on or choose something from the recommended tools here.

I have postponed the laptop/desktop updates in WSUS for a week or so until this stuff is fixed. I don't want to mess with colleagues that complain they are not able to search and find something anymore

There was an issue with US NTP servers multiple users reported during this week. Not sure if it might be the case here but still.

We are a mixed shop, having Mac, Windows and different Linux flavors including some weird ones. IP is the only way for us to get everything working.

But even in a pure Windows environment, I would stick to IP. WSD is crappy. WSD + some HP Devices is a noitmare.

As I already mentioned above the goal is to manage virtual machines. Literally. With salt.modules.virt I can do a lot of things with KVM like list_snapshots, migrate, pause, reset and resume the named VM along with setmem, setcpu and other great stuff. The idea is to have a script or possibly a set of scripts or some kind of self-made console that would allow me to manage VMs centrally on both Hyper-V and KVM the same way without having to rewrite each salt.modules.virt command using PowerShell or direct WMI calls. Of course it's doable and will work but seems to be out of SaltStack concept for me.

Recently implemented 2 nodes starwind hybrid cluster https://www.starwindsoftware.com/starwind-hyperconverged-appliance for one of my customers. It came completely pre-configured including hyper-v, failover cluster, tiering, shared storage and VEEAM backup, all in the same boxes and backup instance with VEEAM was set up FT/HA too. Took me only 10 minutes of a joint remote session with their support engineer to join new servers to existing AD and allow customer's IT guy to proceed with the migration from their old hardware (2 x old SuperMicro servers and a SAN). What I really like about this offering is that Starwind guys allow you to bring own licenses and install free hyper-v on bare-metal if it fits which I find quite incredible. Thus the pricing is usually far below the range Nutanix or Scale do.

BTW they give their VSAN software for free. I like using it for HA file servers on old hardware since it's very simple and works pretty great.

I would split those depending on the infrastructure size like:

• Small business with only 2-3 hosts would be completely fine with Hyper-V
• For a larger one with up to 10 hosts I would prefer going with VMware
• For everything larger KVM or XEN. Probably KVM because of larger community and better self-supporting options...

Currently looking at SaltStack as a centralized management tool for my servers. As far as I see SaltStack has great support for KVM (libvirt) that allows one to do almost everything in a virtualized environment based on KVM. Windows capabilities are pretty great for massive deployment and servers/computers management but I can not find something similar to libvirt capabilities for Hyper-V. The only thing I could find is this one https://github.com/bougie/salt-hyperv-formula but it allows only the installation and configuration of the role itself and is not developed for 3 years already.

So far the only way to manage Hyper-V virtual machines that I see is using PowerShell commands being sent through SaltStack which seems not the best option. Am I missing something?

I've tried Guacamole on CentOS and Debian. Both worked fine, but it was a bare-metal install (not docker). Check Guacamole and Tomcat logs to identify if they are running at all and try to see your connection attempts there. If there is nothing it has to be firewall/connectivity issue.

I've had a similar situation with one of my customers already a while ago. At the end of the day, I landed up with pre-built hyper-converged appliances (a pair of them). Since the budget was the main constraint and DELL was a requirement, I've purchased very customized servers by starwind https://www.starwindsoftware.com/starwind-hyperconverged-appliance (almost no storage and powerful CPU for SQL). These were exactly DELL R630 (not sure if they still ship those) delivered by xByte. One is located in the server room on ground floor and the other one is 2 floors higher. Interconnection is 2 x 10 GBe links. Pretty happy so far.

My primary home lab (2 x SuperMicro 1U servers) runs inside IKEA shelf (slightly modified though). The second home lab (2 x SuperMicro SuperServers E300) is hosted inside LEGO chassis we've built together with my son

Currently, I am in the middle of Patrick Rothfuss - The Name of the Wind.

OK guys, so now we have an option (theoretically) to ship 50 tons of stuff to the orbit.

If you would have an opportunity to do so, what would you bring 150km high? IT-related, of course

Well, true. But Cisco is a kind of standard for networking equipment nowadays and been those for pretty long time already. What would you recommend as an alternative?

@thwr said in Who is at Fault?:

Someone exposed RDP on the firewall? Are you serious? Put a VPN tunnel in front for remote access.

This! Forwarding sensitive stuff like RDP to WAN is just... you know. You can try doing this, however, to see how thousands of brute connections (mostly Chinese IPs) start to initiate within a couple of minutes. Looks pretty scary

Is the "View email messages by conversation" not what you are looking for? It's not exactly the subject-based grouping (meaning everything that has the same subject goes to a single tree) but very similar to it:
https://support.office.com/en-us/article/view-email-messages-by-conversation-0eeec76c-f59b-4834-98e6-05cfdfa9fb07

Currently using ZenDesk (over 1 year already) to support my customers (a bunch of different companies). Pretty simple and convenient interface and nice mobile app (i am using the one for android).

Takes me around 50$/per IT guy/per month (4 engineers in total). Alerting, SLAs and most bells and whistles one might need - pretty happy so far.

You can try using "Project Honolulu" for your management purposes. It's almost a "single pain in the ass" management console that has nothing to do with management itself but rather a configuration tool (basically various MMC snap-ins, Control Panel, Server Manager and so on brought to web) for windows. But it will solve your issues at current stage for sure.
More information here: https://docs.microsoft.com/en-us/windows-server/manage/honolulu/honolulu
Download: https://www.microsoft.com/en-us/evalcenter/evaluate-windows-server-honolulu

A 3rd-party alternative might be the 5nine manager (looks horrible but does it's job) https://www.5nine.com/5nine-manager-standard-edition/ or StarWind Manager https://www.starwindsoftware.com/starwind-manager which is still in preview state or this one http://hv-manager.org/.

@scottalanmiller said in I can't even:

I was actually consulting on a backup design just yesterday where we were using Hyper-V in exactly this way as a backup target, but building the backup system using Starwind. It was a backup device in every way, no expectation of VMs to run there, no live systems ever, just Hyper-V + Starwind used to handle the replica-based file backups.

That's a good setup if you need HA/FT backups and there is a particular requirement to always be able to recover. We have a similar configuration using free version https://www.starwindsoftware.com/starwind-virtual-san-free on top of two decommissioned DELL servers packed with drives and SMB share as VEEAM repository on top. Works like a charm.

A separate SAN or NAS for an SMB deployment? It's an overkill. And a SPoF. And a lot of other things. I would go with two clustered servers and DAS in each. Since it's Hyper-V - you can use S2D https://docs.microsoft.com/en-us/windows-server/storage/storage-spaces/storage-spaces-direct-overview to convert DAS into a single shared storage pool for high availability, fault-tolerance and failover. If you are aware that S2D sucks on less than 4 hosts - go with StarWind https://www.starwindsoftware.com/starwind-virtual-san-free (installs directly on bare-metal Hyper-V, works great on only two hosts) that does the same as S2D and is free. VEEAM for backups. That's it

This tip helped me once with troubleshooting in a similar situation:
https://blogs.iis.net/rickbarber/working-past-500-internal-server-error
Give it a try and tell us how it goes. Hope it helps.