seanho
seanho
Comments
-
Ok then: * Best LES provider: IonSwitch * Best non-LES provider: OVH
-
* IonSwitch 2. DRServer mod edit: fail.
-
Enterprise pull equipment can be so cheap. Instead of paying for a support contract with same-day replacement, just buy two or three spares! (Yes, I know corporate procurement doesn't work like that. But in a SME situation where you can directly make purchasing and support decisions, the cost savings can be tremendous.)
-
I understand, my point is that these old boxes are so cheap nowadays that you can pick one up for a dedicated firewall and put your M4 to work with a hypervisor. You can have a second failover instance of PFSense (with CARP) on a VM if you like, but always have a dedicated box for firewall.
-
And after you learn iptables, now you can learn the "new hotness" that is nftables....
-
R210ii, HP T620plus, T730, M73/M93p SFF, X9SCM/L in a little 1U, lots of options. These are all old tech, but will do very well as dedicated PFSense boxes. For the SFF and thin clients, make sure there's space to add a $20 PCIe NIC.
-
The cutover to nftables default is such a pain. Kubernetes also requires reverting to iptables-legacy.
-
Do you need it to be new with a support contract? If not, used Dell 5324 is dirt cheap and has everything you list except for 10G uplink. To add 10G uplink, Aruba S2500-24T. There are lots of these sort of EOL managed gigabit switches around as enterprise pulls. The S2500 has two SFP+ ports for uplink and two for stacking,…
-
For debian, in addition to what you've listed: git, rsync, locales, apt-transport-https. Perhaps bash, but dash may well be sufficient. Nowadays, just ssh and apt are enough for ansible to install everything else I'd need.
-
This would be awesome! I never considered that it might actually work.
-
Yep, that was my conclusion regarding gluster, as well. I have my doubts ceph would tolerate long latency, either. Maybe it could work if I segment VPSes into geo-regions of latency less than 40ms.
-
I enforce that all server and app config is managed and tracked in git, and all data is in storage buckets that can easily be backed up with common tools like borg. If a server dies, reinstall the OS, hit it with ansible, and restore data from backups. Next step is resilient distributed storage, I'm thinking Ceph or…
-
In the past year, I've started getting into home labbing and now have around 70c/140t in a 36U rack, plus some pretty nifty switches. I thought that would decrease my annual spend on VPSes, but somehow it has only increased....
-
Oh, it's like the old quarterly Top Provider contests! * Kimsufi: I'm sorry, I know OVH is the behemoth everyone loves to hate, but I have an i5 that has been good to me, and I will never ever let it go (as long as they don't raise the price too much). * Inception: I don't have much use for EU presence, but a BF deal on…
-
I admire the effort you put into tracking CC's various shell hosts, and I am indignant at their blatant scamming, but at this point I am just tired of the drama on the old forum, even from a schadenfreude kind of view.
-
I didn't get many new servers this time, mostly just renewed BF deals from previous years. I'm still undecided whether it was better to renew my HostHatch 2.5TB in LA or switch to a 3TB in Chicago. I tracked VirMach's algorithmic deals for a bit, but was not tempted. I probably should have taken advantage of Anthony's…
-
Yeah, that was my plan, too -- consolidate over a dozen small VPSes into one nice dedi. Now I still have a dozen idling VPSes, plus multiple idling dedis....
-
Doesn't AlphaVPS use CC for their LA location? So you'd put Alex on your blacklist, too, even his UK/Clouvider location?
-
Also, just to note that VirMach is not exclusively CC, e.g., at their Seattle location.
-
My small VirMach idlers (mostly just for VPN) have been very stable, probably because they have such strict processes in place for abusers. Consequently, I don't use them for anything remotely CPU-intensive. You would have been nuts not to have gone for that $4/yr 1GB, though, if their billing site hadn't crashed and…
-
I have less need for EU VPSes nowadays, but in the past had only good experiences with LiteServer. Excellent support.
-
Very nice offers; this is too good to pass up!
-
LXC is fine for your own private use, where you control the containers as well as the host node. It is insufficient for VPS providers, where the containers are untrusted. Software RAID can be just fine, but with VPSes the storage is managed by the provider, whether it's hardware RAID or software is transparent to you as a…
-
I lock down most of my fleet to whitelist incoming ssh only from select jump hosts. Ansible helps keep the attack surface small, no new ports opened besides ssh. I have played around with puppet, chef, and salt stack, though; much respect to them.
-
https://www.ansible.com/
-
Ooh, just in time for some nice BF storage deals in Chicago!
-
I think a hidden, LES-only WHMCS/Blesta category would make the scraping opt-in for providers. No need to manage codes per provider, then.
-
I guess maybe 28k. Thanks for a bit of fun, although I wouldn't need another OVZ!
-
Great offer! Would I be able to request a smaller SSD instead of the 500GB HDD?