New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Comments
I am. Opened a ticket last night to check so waiting to hear on that. Plan was a storage node for LA where they had the issues the other week so I have been assuming there’s a delay as a result of that. They did provision a compute node, also for LA, pretty quickly but
Also LA, maybe they are still having issues.
They seem to have significant cpu steal on every server. I tested it a year ago and measured cpu steal of 11%. I tried again a week ago and getting 2-3% steal on a server in hong kong with nothing running.
Support responded and the orders haven’t dropped off the radar. They’re making sure the earlier issues don’t arise again and also getting some additional kit.
How would you test for that?
Steal? You type top and look at the number for st.
I don't have any servers in Asia, but I have never seen any noticeable steal with HostHatch.
This might be related to various cryptocurrency mining activities, and it's taking time for HostHatch to detect and kick off the abusers.
I tested chicago too, it also had significant cpu steal.
Yes.
It's worse for me (around 10% CPU steal, sustaining
).
2 VPS in LA are ok.
In the SG server, I'm seeing 3-7% st.
Yup, me too performance issue on Singapore
Anyone managed to boot up from ISO?
i have tried monting but keep stuck in offline mode
you are not alone, I'm waiting too, 28.May make the order for HK storage
@skippy you use almost 0 CPU..
That is a sign of being a true LET idler. Can't have CPU steal visible if you aren't using any CPU
Awesome. Now that is a low-end idler. Congratulations.
Second @aj_potc 's comment that until recently I've rarely ever seen any significant steal either. Whatever this latest crypto is though had one of my Chicago services at ~30% sustained steal for several days. Looks like they finally caught whomever it was this morning may they rot in hell.
We've seen a recent hike in abuse, that's (largely) been taken care of, and should be completely taken care of in the next 24-48 hours.
RE: HK storage VMs, these should be delivered in the next 24 hours.
Sorry for the delays on these things - we've been swamped with multiple issues (with some customer(s) continuing to create hundreds of accounts with stolen CCs and somehow passing through our + stripe checks). This should be totally under control in the next couple of days.
LA Storage:
Intel(R) Xeon(R) Gold 6140 is good. (Finally no more E5-2xxx.)
How about the new panel?
Basically a wash with my E5-2690v4 node
I think it’s a golden example of the previous gen storage offerings though
@hosthatch Any update on the availability of the firewall feature?
@hosthatch JP node down?
If a node is having issues that are going to take a look time to resolve:
If a node is having issues that we can resolve faster (such as a reboot), you will receive an email, and we will keep you updated in tickets.
The new panel is delayed by 2-3 months, but we'll bring both bandwidth sharing and firewall in BETA access in the current panel. I'm afraid I don't have an ETA yet, but sometime soon.
I assume you opened a ticket, and we resolved it pretty fast? "Steal" is a completely valid complaint, and is assigned one of the highest priorities in support. We resolve it as soon we are made aware
I tried again a week ago and getting 2-3% steal on a server in Hong Kong with nothing running.
We don't investigate complaints of < 3% steal, because you are running on a virtualized server. If you see any significant steal when you do use more CPU, please open a ticket and we'll fix it right now. It should never happen in the first place.
We're not doing the "25 dedicated cores where everyone is competing for the same cores" and the steal is considered normal. I wrote the following text which will go in our docs at some point.
The young troll that you see that opened a thread about how HostHatch is selling fruit and not CPU should be considered a troll, considering the thread came with zero documented of actual fruit selling. Please be rest assured that CPU steal (at any significant level < 3%) is not considered acceptable at HostHatch.
The only acceptable steal is when we apply CPU limits to a VM, and that shows up as steal inside the server (ie you have 4 dedicated cores and 8 total cores, and you are using 800% CPU consistently, you will be limited to 4 dedicated cores, and that will show as high steal inside the VM because it still technically has 8 cores)
No it wasn't fixed, I made a ticket for it and it wasn't responded to so I just let it expire.
If the server has nothing running, a steal of that high is honestly really high and it indicates that it would be much higher if it wasn't idle.
It’s funny that you mention ’thread’, because that’s precisely what you adverise as ’cores’, you lying creepy old man.
They were probably busy reading my thread.
For what it is worth, I am still their customer and using them for applications that can tolerate a moderate amount of CPU steal. It seems to be better now but maybe I just lucked out and got on a good node. I'm also not monitoring the cpu steal in all the time anymore so it could be higher.