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
Just had a look at my @Chunkserve VPS. It's up and running fine but was rebooted about 7 hours ago.
I don't concur with OP's assertion. Granted, I had only very limited (need for) interaction with CS's support but neither had problems with it nor to wait long for a response.
Also: This is a ridiculously dirt-cheap VPS, so of bloody course one can't expect super-fast and high end support. But, at least in my experience, Chunkserve gives you what can reasonably expected in terms of support.
Well, actually IMO that's the right order. After all, getting one's server up again is more important than some words (ticket reply)...
But of course a first reaction (and looking at the problem asap) is important.
Yes and no. If you can prevent people creating tickets because you put a message of downtime somewhere in a obvious place on your portal/site (e.g. just before a ticket can be created!) then the number of tickets issued will be much less and at the same time you display to be "in control".
If I want to issue a ticket @09:22 about outage started @08:36, I wouldn't create a ticket and I know that the issue is being looked at.
If I don't see it, don't get a ticket reply, see others react on LET that there is an issue and then around 12:15 the VPS returns it can also be that there was a simple reboot that was issues at around 12. And that nothing has been done between 08:36 and 12:00.
And yes, I understand this is LET with cheap services. But "Enterprise-grade hardware,
high uptime and proffesional support are just few of our advantages" does IMHO not match the experience of "regular downtime" and "hours or days before a node is restarted" (apart from spelling mistake in professional). And looking at comments of others I have been quite lucky.
I have 2 cloud services with Chunkserve since November 2024. I barely had any issue with them. I only opened 2 tickets and those were handled quite fast.
Keep it going with the work and keep improving @Chunkserve
Not trying to avoid anything. We started diagnosing the cause and fixing everything before 10. I noticed around around 8:30 a ~2minute ICMP probe fail + TCP unsuccessful probe to pve, but everything went up after that. Today, we made another probes for cluster-related issues - node didn't get the quorum, so every action failed. - so such things will be visible in the future. My first message on LET today was issued when it went up and bulk start task was running. Post appeared after most of the services went up - but the op didn't message us about disruption - just received telegram message with post link - then I opened it and issued my statement. Regarding previous indicents - especially from jan-feb.2025 - some of them were related to the transit provider (we're multihomed now), some of them to the core stack (we moved to new Cisco one) - and the network itself didn't have any major downtime. I'm not enabling monitoring notifications from every vm - tht would make no sense. Please, point me out where I avoided Your question - "but it doesn't help when your answers are avoiding these questions". I'm trying to be as transparent as possible - so explained how my side looks like.
I can partially agree. Even if there was any planned downtime/incident, and there was a banner on the top of customer area + entry on status page, we're still receiving multiple tickets. Of course, it may be limited, but it won't prevent everyone from creating it.
again, explained how we saw that in the previous message. Every node has ICMP probe for it's IP (both public one and VPN range) + Virtualizor TCP + Proxmox TCP probe + SSH. I added today alerting about service fails (pveproxy, pve-cluster and corosync). After the anomaly ~8:30, everything was green on our status dashboards - and that leaded to a wrong assumption. We received couple of tickets around ~9-9:30, and then I investigated it further and started diagnosing+fixing the issue. Of course, it should not look like that - but everyone learns from mistakes - and if I want to be as transparent as possible, I'll let You know about it. (not trying to avoid anything). Server wasn't rebooted around 12. Before 12, cluster was fully functional and I ran Bulk VM start task. After it was processing, I started replying on LET, tickets, TG messages (and around that, op created the thread).
That is also not right - explained it in the previous message. From customer perspective it may look like that, but not every issue may be resolved instantly (as in this case, it took some time to re-establish the cluster, fix the system, ensure the stability and process with restoring VM access)
"And looking at comments of others I have been quite lucky" - may You please clarify? Since some time, after eliminating most of the issues, we're receiving (publicly and privately) mostly positive feedback. I would love to explain the particular situation - just please let me know about it.
As you said: Yes and No. Because humans being humans, chances are that a no matter how obvious message on the provider's site, quite a few will send a ticket anyway.
I get your point, but one also should understand that this happened during a weekend and that, especially smaller, companies often just aren't in fully operational mode on a sunday.
As for the quoted text of CS, I'd take that as something in between marketing bla and a declaration of good intentions.
On sunday, we have limited "workforce" - as not everyone is fully avaliable. That was the reason of delayed reaction (now it won't happen, because monitoring also reports state of specific services - such as pve-cluster, corosync and pveproxy) - so if ie. cluster will be down and service crashes, I'll be informed. I also created a tiny VM on every node, that would be probed by ICMP, to measure also example guest state.
We're not a big company, so our capacity may be limited, but we're working on that - and that lead to delay regarding situation that happened today. Previously, we encountented some problems with support, and now it looks better. At the moment every ticket is answered, and we're trying to keep it like that. Some customers, that bought servers ie. in nov-dec.2024, and encountered such issues in the past, are now giving us positive feedback about improvement - and that is the most important thing.
So, to summarise my previous messages.
I can fully agree, that it should be done quicker - and my 2 colegues were not avaliable in the morning (since it's sunday), I was in a rush organising topics regarding PL infrastructure (planning calls, writing emails, "sketching" infrastructure). Now, as I said multiple times, cluster/proxmox related services are attached to monitoring system too. It's 100% our fail, that it wasn't detected faster (speaking about the ~1hr to start diagnosing everything). We also learned a valuable lesson and already implemented a solution that won't allow it to happen again.
No opinion on chunkyserve or whatever, but until I see some notification that you're aware your server is on fire, I have to keep wasting time trying to notify you. I charge a lot of money for my on-call service!
This is not an uncommon scenario in LowEndLand, hosts that don't even monitor their hardware or connectivity.
What we are discussing here ?
I have some tickets with the support and every ticket was answered.
Yes, sometimes it took a bit longer and yes there could be interruptions.
The support is doing a great job
But you also cannot expect 100% uptime on a $12 product.
If you expect a HA Server go to Microsoft azure or some other big player and pay a big amount.
Edited: read further down.
You mean create a support ticket?
That's a good point. You're not paying a lot of money to @Chunkserve for on-call service.
Dont worry not every LET host is like that
You're still not getting the point: my perception as customer may not be correct but nothing has been done (communication or other signs) that proves otherwise. Benefit of the doubt can be applied only a limited times.
Anyway, what has been done or what has been seen leads to an endless yes-no discussion I think.
If you have an issue I should be able to see it on your status page or you've posted an announcement. If you do neither and I have to put in a ticket I'm going to wonder why I have to notify you that your shit is broke.
The op was complaining that this provider doesn't respond to tickets anyway.
I'm not paying any money at all to chunkyserve, they don't pass my sniff test.
I agree. About the first thing a provider should do is as you said. Only when done and the node(s) is/are up again, tickets are secondary. The important thing is that everything runs fine again.
You're very well fuck sir. Full year fuck with vps.
It might be illegal sir
Problem happens. Check status page. If no entry, create ticket. That's it.
I experience issues with Slack and Onelogin where their status pages are not updated for about 15 minutes after I first experience an issue.
Slack and Onelogin have millions more in resources and monitoring. Your expectation isn't necessarily correct.
I got caught up in a little righteous indignation sorry, but ya status > fix > tickets.
Only vaguely related but boy fake status pages really twist my tail.
Well, it's correct for me
These kids learning how to run a business by carving out vps on a rented node and hawking them on LET don't get a lot of slack from me, there are dozens to choose from and new ones every day.
communication is key
@Chunkserve its a prem hoster with a great support. I have a vps with him and the support was good for the cheap price i paid.
I also believe the same. Maybe they would need to improve their monitoring of services to get faster and better notifications, even before some customers would even dare to think about opening tickets, but I don't know what's behind the scenes already.
And on that note, maybe @HBAndrei or @faleddo can make a special offer to @Chunkserve for low-end monitoring solutions.
For myself i paid €11.99 year vps with 3GB RAM 3vCores and 25GB SSD , i had problems with network outages twice but he fixed everything and ticket was answered by him by other day so for what i paid the service it's good im not judging the OP. But i belive that the @Chunkserve its a good dude and offers good services. The best thing to do it's to communicate , and the monitoring solutions would be great for sure.