Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!


Introducing Singapore - deploy fast, consistently performing NVMe KVM servers - Page 7
New on LowEndTalk? Please Register and read our Community Rules.

Introducing Singapore - deploy fast, consistently performing NVMe KVM servers

1234579

Comments

  • Weird, it's Saturday the 31st here. Is it not Saturday for some of you?

  • hosthatchhosthatch Member, Top Provider

    I was in 3 different timezones, flying for nearly 24 hours over the last weekend, and wrote that email on a plane. So it's totally my fault.

    Sorry for the mistake of typing Saturday the 31st. It was Tuesday the 31st. (Kind of obvious because the email was sent after Saturday...but regardless sorry for the confusion).

    In better news - we will no longer have the network issues people were seeing earlier in Singapore and will have much smoother sailing going forward. :)

    Thanked by 2explicit cybertech
  • brueggusbrueggus Member, IPv6 Advocate

    @hosthatch said: we will no longer have the network issues people were seeing earlier in Singapore

    I'm sorry to disappoint you...

    $ ping6 dns.he.net
    PING dns.he.net(dns.he.net (2001:470:0:149::2)) 56 data bytes
    ^C
    --- dns.he.net ping statistics ---
    17 packets transmitted, 0 received, 100% packet loss, time 15999ms
    
  • brueggusbrueggus Member, IPv6 Advocate
    edited August 31

    @brueggus said:

    @hosthatch said: we will no longer have the network issues people were seeing earlier in Singapore

    I'm sorry to disappoint you...

    $ ping6 dns.he.net
    PING dns.he.net(dns.he.net (2001:470:0:149::2)) 56 data bytes
    ^C
    --- dns.he.net ping statistics ---
    17 packets transmitted, 0 received, 100% packet loss, time 15999ms
    

    Aaaaand it's fixed. 🎉

    @hosthatch said: I was in 3 different timezones, flying for nearly 24 hours over the last weekend

    Hope you'll be able to get some rest now.

  • jon617jon617 Member
    edited August 31

    @hosthatch said: In better news - we will no longer have the network issues people were seeing earlier in Singapore and will have much smoother sailing going forward

    Don't speak too soon! You still have network issues in Singapore. See my ticket!

    I did some traffic inspecting, and I think I figured out your network problems. Also, see my ticket @hosthatch I just shared all this info with you, sharing here too.

    My server has a firewall, as many other do. I wonder if I may have a firewall rule blocking something needed by HostHatch. For me, I noticed an IP on the network hits the broadcast IP, which my server listens to, with a UDP packet. When I allowed all inbound UDP traffic to my server, my network issues went away in Singapore.

    I think the "ping" may come from IP x.x.x.17. I created a rule allowing UDP from that IP, and so far so good. I set my firewall to allow ALL inbound UDP from that IP, rebooted, and so far so good.

    HostHatch's correct fix should be to use arp, dhcp, or something more common to determine which IPs are online, rather than a "ping" every 30 seconds. Or, at least, post an announcement to customers to allow this unusual "ping" setup in Singapore.

    EDIT: This time it was me who spoke too soon. Network problems still happen after firewall rule change. I'll keep looking at packets. One consistent thing is, when I boot with all firewall rules disabled, network seems fine.

    Thanked by 1explicit
  • brueggusbrueggus Member, IPv6 Advocate

    @jon617 said: HH has a UDP "ping" that hits the broadcast IP every 30 seconds, to port 21027. If your server's firewall blocks this UDP traffic, your server will drop off the network.

    I'm pretty sure what you're seeing is Syncthing discovery broadcast traffic. You should see tons of ICMP and ARP traffic if you run tcpdump for a while. Hosthatch's not doing anything fancy here.

    My server used to lose its network connectivity due to their router sending me ICMP Redirects to routers which weren't on link. I set sysctl parameters to ignore them and my IPv4 network is stable since then. I've no idea if that issue still exists.. I just left everything as-is as soon as it worked.

    Thanked by 1vimalware
  • jon617jon617 Member
    edited August 31

    @brueggus said: I'm pretty sure what you're seeing is Syncthing discovery broadcast traffic. You should see tons of ICMP and ARP traffic if you run tcpdump for a while. Hosthatch's not doing anything fancy here.

    Yeah I realized that after making this post... But I have noticed consistently when I turn off my firewall, then reboot, all networking is fine. Made me think there was something that breaks network when my server blocks certain traffic, and the UDP from x.x.x.17 was an obvious candidate since I saw it every 30 seconds.

    To summarize where my troubleshooting is at. When I change my firewall to allow all inbound and outbound traffic, then reboot, all networking is fine. This is also consistent with why networking is fine right after a fresh install of the OS, when no firewall rules are in place. When I slowly re-enable my firewall rules, the network does soon die. If it's not caused by the UDP from x.x.x.17 getting no answer, i'm still curious if there's another firewall block that would cause their network to kill routing for my server.

  • jon617jon617 Member
    edited August 31

    @brueggus said: My server used to lose its network connectivity due to their router sending me ICMP Redirects to routers which weren't on link. I set sysctl parameters to ignore them and my IPv4 network is stable since then. I've no idea if that issue still exists.. I just left everything as-is as soon as it worked.

    What sysctl.conf settings did you use to ignore icmp redirects? Something like:

    net.ipv4.conf.all.accept_redirects = 0
    net.ipv4.conf.default.accept_redirects = 0 
    net.ipv4.conf.all.send_redirects = 0
    

    I will try this as well.

    Thanked by 1vimalware
  • @Panit said:

    @brueggus said:

    @Panit said: I did its. But It can't fix.
    I try to reinstall form ISO but still has the same problem.

    What does lsblkshow for you?

    lsblk

    In the future, you can try googling for tips or looking at the result of what you were asked to run. And also what made you think you had a 1.8GB partition? You had a 20GB partition as shown in your screenshot. Giving bad information will only make troubleshooting problems take longer.

    Thanked by 1webcraft
  • @TimboJones said: You had a 20GB partition as shown in your screenshot

    Technically we only know the size of that LVM volume. The filesystem on it could be significantly smaller if whatever resized it didn't do the filesystem at the same time.

  • K4Y5K4Y5 Member

    @hosthatch Still having issues with my SGP VM, same as day 1. Not able to SSH into my VM or PING the IP as it times out. However, Hetrixtools shows that the instance is online. What gives?

    Reboot temporarily fixes it. But that doesn't last more than a few minutes.

  • PanitPanit Member
    edited September 1

    @TimboJones said:

    @Panit said:

    @brueggus said:

    @Panit said: I did its. But It can't fix.
    I try to reinstall form ISO but still has the same problem.

    What does lsblkshow for you?

    lsblk

    In the future, you can try googling for tips or looking at the result of what you were asked to run. And also what made you think you had a 1.8GB partition? You had a 20GB partition as shown in your screenshot. Giving bad information will only make troubleshooting problems take longer.

    Sorry for my tiny knowledge. English isn't my native language.
    And I'm just an end-user. I have inexperienced. And I used to use the service of Rackneard, Virmarch, and some provider in LET. All of them never has this problem. I just used their control panel to install the OS and all ready to work.
    This is my first time that after reinstall OS form the provider template then I must to find the way to fixed its.

    I know this is value VPS I can't hope for best service. So I waiting for the support team to answer my ticket but it more than 48 hours before I posted In this tread because I saw the other user got the same problem too.

    I'm sorry if my inexperienced made problems.

  • any EU offers?

  • DPDP Member, The Domain Guy

    @mad_4u said:
    any EU offers?

    Probably in 2 months.

  • jon617jon617 Member
    edited September 2

    Emil from HH updated my ticket today, saying the ICMP redirects should stop now. I can confirm no more ICMP redirect messages appear on the console, network has been online, and I'll keep an eye on my server's console for a few days to be sure. Hopefully the issues are no more in Singapore, but if anyone else's Singapore server is still unreachable over the network, try stopping and re-starting your server.

    If you do see an error in the console saying "redirect host" and "new nexthop", try adding these to your sysctl.conf then reboot

    net.ipv4.conf.all.accept_redirects = 0
    net.ipv4.conf.default.accept_redirects = 0 
    net.ipv4.conf.all.send_redirects = 0
    

    (thanks @brueggus for the sysctl tip)

  • @Panit said:

    @TimboJones said:

    @Panit said:

    @brueggus said:

    @Panit said: I did its. But It can't fix.
    I try to reinstall form ISO but still has the same problem.

    What does lsblkshow for you?

    lsblk

    In the future, you can try googling for tips or looking at the result of what you were asked to run. And also what made you think you had a 1.8GB partition? You had a 20GB partition as shown in your screenshot. Giving bad information will only make troubleshooting problems take longer.

    Sorry for my tiny knowledge. English isn't my native language.
    And I'm just an end-user. I have inexperienced. And I used to use the service of Rackneard, Virmarch, and some provider in LET. All of them never has this problem. I just used their control panel to install the OS and all ready to work.
    This is my first time that after reinstall OS form the provider template then I must to find the way to fixed its.

    I know this is value VPS I can't hope for best service. So I waiting for the support team to answer my ticket but it more than 48 hours before I posted In this tread because I saw the other user got the same problem too.

    I'm sorry if my inexperienced made problems.

    No need to be sorry, just saying while waiting hours and hours for a support reply, you can search and expand your knowledge. That's just general advice that applies near universally in the age of Google.

    Thanked by 1chocolateshirt
  • Hello, I have purchased your service,

    Invoice #245529

    Can you automatically get double ram, double storage and double bandwidth?

  • @xiangge said:
    Hello, I have purchased your service,

    Invoice #245529

    Can you automatically get double ram, double storage and double bandwidth?

    Thank you for your patiente. We've doubled your ram, storage, bandwidth, and IPv6.

    Thanked by 1giang
  • It seems they have stocks now :D

  • explicitexplicit Member
    edited September 2

    @yoursunny said:

    @explicit said:
    Downtime shouldn't be happening on Singapore timezone where customers are running production critical workloads on it.

    Singapore timezone covers all 24 hours a day.
    If the maintenance cannot occur within Singapore timezone, is it suppose to wait until the end of time?

    You are noticed in advance regarding an upcoming downtime.
    If you have production critical workloads, you should reroute traffic to your other servers ahead of time.

    whoops, Hi, thanks for your reply.

    I assume you weren't sarcastic in your response, but i'll take it genuinely that you have spent time understanding why an downtime was issued less than a 3 days without further details. Moreover there was a mistake in the scheduled maintenance day & date - which was futher clarified by HostHatch. What I meant was Singapore office hours (8am - 6pm) instead of timezone.

    I will consider rerouting the traffic in such case and consider a LB/proxy solution down the road.

    None the less, HostHatch has been stable thusfar for their Singapore servers and i am still pleased with it.

    @Faizi said:
    Clearly, they made a typo. It should be Tuesday 31 August 2021.

    It was down just now exactly at 8:00p.m (GMT +8), and up again at 9:00p.m.

    Why are you running a 'production critical workloads' on a new launch product from a budget LET provider?

    This workload is critical, but still small in size that doesn't necessitate a dedicated hardware which i've moved out of into HostHatch.

    I'm relatively new to LET, hope to get some advise. Should I avoid newly launch product infuture? And is HostHatch a budget LET provider?

    Thanked by 2yoursunny Faizi
  • @jon617 said: still

    @hosthatch said:
    I was in 3 different timezones, flying for nearly 24 hours over the last weekend, and wrote that email on a plane. So it's totally my fault.

    Sorry for the mistake of typing Saturday the 31st. It was Tuesday the 31st. (Kind of obvious because the email was sent after Saturday...but regardless sorry for the confusion).

    Thanks mate!

  • @explicit said:

    I'm relatively new to LET, hope to get some advise. Should I avoid newly launch product infuture? And is HostHatch a budget LET provider?

    No. You grab it before it's sold out and let it idle.

    Every provider on LET is a budget provider.

    Thanked by 1webcraft
  • @hosthatch said:

    > 1 CPU core (12.5% dedicated, burstable up to 100%)
    > 1 GB RAM
    > 10 GB RAID-10 NVMe
    > 500 GB bandwidth
    > $20 per year
    > Pay for two years - get doubled RAM, storage, and bandwidth
    > https://manage.hosthatch.com/billing/order/sgp-1g
    > 
    > 2 CPU cores (50% dedicated, burstable up to 200%)
    > 4 GB RAM
    > 20 GB NVMe SSD
    > 1 TB bandwidth
    > $35 per year
    > Pay for two years - get doubled RAM, storage, and bandwidth
    > https://manage.hosthatch.com/billing/order/sgp-4g
    > 

    what differences between RAID-10 NVMe and NVMe SSD ?

  • edited September 3

    @astrovps said:
    what differences between RAID-10 NVMe and NVMe SSD ?

    All package use RAID-10 NVMe.

    NVMe SSD is a SSD that utilizing NVMe interface. SATA SSD is a SSD that utilizing SATA interface. The former one have higher data transfer, more than 6 Gbps.

  • ChandraChandra Member
    edited September 3

    Picked a 4GB VPS two and half hours back, my first with hosthatch (Inv#245623). Now waiting for provisioning and doubling resources. :)

    Does it usually take so long?
    Thanks.

  • @Chandra said:
    Picked a 4GB VPS two and half hours back, my first with hosthatch (Inv#245623). Now waiting for provisioning and doubling resources. :)

    Does it usually take so long?
    Thanks.

    Congratulations on your first post.

  • @Chandra said:
    Picked a 4GB VPS two and half hours back, my first with hosthatch (Inv#245623). Now waiting for provisioning and doubling resources. :)

    Does it usually take so long?
    Thanks.

    You must be new here.

    *checks profile*

    Yep.

  • ChandraChandra Member
    edited September 3

    Thanks @dosai
    A long timer on LEB but realized I never posted on LET or don't remember if I posted from another email!

    Spoilt by instant VPS set up emails that come even before the payment receipts come, I guess :D

  • BHSBHS Member

    Hi, @hosthatch

    Invoice #245708 please doubled RAM, storage, and bandwidth.
    Thanks

  • Why is there an expectation that resources will still be doubled when the sale was for 24 hours and ended days ago?

Sign In or Register to comment.