Brandon from eNetSouth has just put up some new servers at SoftLayer in San Jose CA, and asked me to “catch us off guard”. So here’s the deal from them. With this sign up link, $7/month gets you this:
- 2GB memory
- 50GB storage
- 1500GB/month data transfer on 1Gbps
- 1x IPv4 + 1x IPv6
- OpenVZ/SolousVM
Those are powered by Xeon E3-1270 Sandy Bridge CPUs. Test IP: 50.23.74.242. Instant activation, subject to IPv4 availability from upstream.
eNetSouth has grown quite a bit since Brandon started working full time on it. Now they have Brain working as full-time support there and Miguel doing part-time between colleges, so hopefully the ticket turn-around time would be shortened.
Disclaimer: LowEndTalk, the community website, runs on a VMWare server sponsored by eNetSouth.
Featured Offers

Have you benchmarked your server yet?
Run a free ServerVerify benchmark and compare real-world CPU, disk and network performance with community-tested VPS and dedicated servers.
Related Posts:

ColoCrossing's Dedicated Server Sale Features Systems as Low as $25/Month! Los Angeles and Buffalo!

Host4Fun: 50% Off in Over 20 Locations! Get an 8GB RAM VPS for Only $5/Month!

Happy Valentine's Day from ColoCrossing! Get a VPS for Only $1.97/Month Plus Tons of Deals on Cheap...

LAST DAY to Win a Ryzen 5 Laptop, Govee Floor Lamp, Cool Socks, and More from RackNerd! And a 1GB VP...

Last Chance for RackNerd Black Friday! Plus a Bonus Code for Their January Giveaway! Get a 1GB VPS...

well,if the server is fine now:)
when the tun thing can be resolved?
am i asking for too much?lol
Not too much. Lets make sure this is stable first.
hope this time it would be stable:)
and if the tun can be enabled then it would be perfect.lol
Yeeey, server back online.
Lets see how this latest change works out. Thank you all very much for your courtesy, patience and understanding while we get this worked out. You don’t know how much I appreciate it.
I’m not one for taking “stabs in the dark” to find a resolution.
I notice there is a upstream patch for the current kernel I am running. I’m going to see if it has a fix for what we are seeing or give up on 6.0 2.6.32 and fall back to 5.6 2.6.18
One of those two things are happening today.
Brandon
I have successfully ordered three days ago, I hope will be long-term service
Brandon could you help me out with ticket #117198 ?
Thanks
Update: ovz6 is performing perfectly at 2.6.32 kernel/ 042stab037.1 openVZ after the last changes.
Enjoy! Please contact support if there are any outstanding issues.
what about tun? does it mean you had solutions for this?>
VPS Support 09/21/2011 00:36 Mail System Issue? Open Medium
subbed
My VPS is still unusable… >.<
Solus marks it as offline, SSH hangs after I type in my username before I key in my password.
ya, it is off.
down again? load average more than 12
yeah it down again…
brandon is busyman since yesterday…
out of stock.
What happens when the new stock?
This is totally not fair, not at all.
I wouldn’t complain, $7 for a 2GB VPS, and Brandon is working on CentOS 6 with all that kernel trouble.
If I had the $7 to spare (I don’t… really) I’d totally chime in. 😛
Hey, at least he’s on here, telling you whats going on, and at least keeping everyone updated. Half of the providers barely do that.
:< donw again.
This is Brian posting under Brandons account because of the validation.
There is another problem. I have the node back up and looking at logs to find out why it went down again.
I will start the containters as soon as I know.
My server is back online! Solusvm shows it as online, and SSH access is fully restored. The previous random hangs is now gone too it seems. Wonder if this means everything’s solved, hopefully not just another respite. =)
When were you seeing these random hangs? Previously this evening or all day today or all the time prior to this last issue?
(P.S.. trying to use the internet with 300+ms latency is sad.. very sad)
Are you sticking with 6 or thinking moving to 5.6?
Brandon do you have some HTTP caching program running on port 80? I have a non-HTTP server that ran on port 80 with my old host fine, but here using telnet on port 80 from my home(note: telnet 127.0.0.1 doesn’t seem to give this problem) I get a HTTP response out of nowhere.
I’m deciding on trying the latest 2.6.32-042stab039.2 or going back to 5.6. Going back to 5.6 would be the longest downtime but almost a certain fix.
Going to 2.6.32-042stab039.2 is the last item left in the bag of tricks. I’m beginning to thing that sadly, 6 is simply junk at this point.
Read below- I replied on the wrong line.. (beyond tired at this point.. )
Yea Me too
But something has been change
The ram size changed to 1.95GB from 2GB 🙁
My VPS was previously offline most of the time I tried to use it, but when it was online it kept hanging. But it seems since my last reply all has been fine.
I replied wrong to you when answering John. Read above.. Anyways – there is nothing that I would have running on your IP on port 80.
These bugs are what I’m seeing as well. No word if 39.2 resolved them.
http://bugzilla.openvz.org/show_bug.cgi?id=2027
http://bugzilla.openvz.org/show_bug.cgi?id=2015
http://bugzilla.openvz.org/show_bug.cgi?id=2022
Performance is much better today than it was on Friday. Current uptime is 10 hours.
@Brandon: Hate to use LEB as a support desk, but is there any chance you can take a look at [Ticket ID: 450756] please? Thanks!
Things are appearing good:
top – 05:15:47 up 15:14, 1 user, load average: 0.00, 0.00, 0.00
That’s the load on your personal VPS and doesn’t reflect what other containers in the same node are doing. It would be interesting if the host allowed users to monitor the overall load on the entire node.
People might get shocked if they were able to do that. Maybe ramhost is the only vendor who provides real time node monitor for OpenVZ? Too bad I’ve never had a chance to grab one plan from there.
On (some?) Linux-VServer VPSes you get to see whole node CPU loads 😀
And data transferred seen in ifconfig eth0 is for the h/w node too 🙂
+1 to rm. I have used Dreamhost VPSs and have that behavior.
Believe me, sometimes it reflect it. When you are doing nothing and have a load of 5 then you have problems…
But I guess this post was to show the uptime.
I wasn’t showing top for the load, I was showing it for the uptime…. -.-
The fix appears to be this:
Linux ovz6.enetsouth.com 2.6.32-042stab039.2 #1 SMP Fri Oct 7 13:21:15 MSD 2011 x86_64 x86_64 x86_64 GNU/Linux
Compiled last Friday. Patched around 4:50 this AM.
Btw @Brandon, the RAM decreased to 2000MB, but shouldn’t 2GB be 2048MB?
Not that it matters, I’d hardly cross the 128MB mark anyways.
During one of the crash events over the past two days SolusVM stopped “accounting” for containers. It was 100% everything for everyone. A simple script was applied to the node to set accounting for the containters. Why it set 2000 instead of 2048 I do not know. Also it seems some containers were set to 2 cpu instead of 4. The values are defined at the top of the script and applied the same to everyone. These can be set back now in SolusVM which was not possible yesterday.
Are we supposed to be allocated 2, or 4 cores?
hope this time will be stable….
yes.. the vps performance is much better… seem brandon move it to another hardware…
4
If you have not already opened a ticket please verify and open one requesting for it to be changed.
Hey Brian, would you take care of Ticket #252901.?
thanks, it was fast.
Regarding the CPU cores, could you guys look at ticket #102329?
how is the tun thing now?
didn’t got a reply of my ticket
same here
I can’t connect to my server, submitted a support ticket 5 days ago, still no answer…
vps is down again:<
My VPS is up right now, but my minecraft world is currupt again.
my is up too..but it’s the 2nd or 3rd reboot in 24hrs
We have decided to give up any hopes of using 6.0 kernels.
We are going back to 5.6 today. There will be serveral minutes of dowtime today while the containers are stopped and the OS is rolled back.
It seems without a doubt that this will resolve the crashing and lockups.
yep,make it more stable first..
i don’t see anything fancy of 6.0 but the endless crash:<
and if the 5.6 works fine,please try to solve the tun thing.i really need my openvpn.lol
What about replying me? 🙂
so all of this will finish today?
A dumb question, is it due to the kernel or CentOS? Personally I don’t like CentOS much because of its delay and lack of support. If possible you may try SL on host instead.
But as I said it might be a dumb question. I’m not an admin.
Another dumb question, what means SL. Softlayer? lol
Scientific Linux 😛
It is being considered as a less invasive approach to resolution. There are discussions out there that it resolves this exact problem:
Oct 10 02:15:20 ovz6 kernel: [76545.975075] [] ?
vzdq_aquotq_lookup+0xf7/0x150 [vzdquota]
Full trace below:
Oct 10 02:15:19 ovz6 kernel: [76545.390522] WARNING: at fs/inode.c:792
unlock_new_inode+0x62/0x70() (Not tainted)
Oct 10 02:15:19 ovz6 kernel: [76545.435410] Hardware name: X9SCI/X9SCA
Oct 10 02:15:20 ovz6 kernel: [76545.457928] Modules linked in: vzethdev simfs
vzrst vzcpt nfs lockd fscache nfs_acl auth_rpcgss sunrpc vzdquota ip6t_REJECT
ip6table_mangle ip$
Oct 10 02:15:20 ovz6 kernel: [76545.813553] Pid: 130726, comm: cat Not tainted
2.6.32-042stab039.2 #1
Oct 10 02:15:20 ovz6 kernel: [76545.852156] Call Trace:
Oct 10 02:15:20 ovz6 kernel: [76545.866915] [] ?
warn_slowpath_common+0x87/0xc0
Oct 10 02:15:20 ovz6 kernel: [76545.904083] [] ?
warn_slowpath_null+0x1a/0x20
Oct 10 02:15:20 ovz6 kernel: [76545.940086] [] ?
unlock_new_inode+0x62/0x70
Oct 10 02:15:20 ovz6 kernel: [76545.975075] [] ?
vzdq_aquotq_lookup+0xf7/0x150 [vzdquota]
Oct 10 02:15:20 ovz6 kernel: [76546.017363] [] ?
do_lookup+0x1c6/0x250
Oct 10 02:15:20 ovz6 kernel: [76546.049735] [] ?
__link_path_walk+0x89e/0x1110
Oct 10 02:15:20 ovz6 kernel: [76546.086263] [] ?
__d_lookup+0xad/0x150
Oct 10 02:15:20 ovz6 kernel: [76546.118640] [] ?
__link_path_walk+0xe68/0x1110
Oct 10 02:15:20 ovz6 kernel: [76546.155170] [] ?
path_walk+0x6a/0xe0
Oct 10 02:15:20 ovz6 kernel: [76546.186542] [] ?
do_path_lookup+0x5b/0xa0
Oct 10 02:15:20 ovz6 kernel: [76546.220559] [] ?
do_filp_open+0xfb/0xcd0
Oct 10 02:15:20 ovz6 kernel: [76546.254050] [] ?
call_filldir+0x7c/0x120 [ext4]
Oct 10 02:15:20 ovz6 kernel: [76546.291157] [] ?
__utrace_engine_release+0x0/0x50
Oct 10 02:15:20 ovz6 kernel: [76546.329207] [] ?
__mutex_lock_killable_slowpath+0x70/0x1d0
Oct 10 02:15:20 ovz6 kernel: [76546.371965] [] ?
strncpy_from_user+0x4a/0x90
Oct 10 02:15:20 ovz6 kernel: [76546.407516] [] ?
alloc_fd+0x92/0x160
Oct 10 02:15:20 ovz6 kernel: [76546.438819] [] ?
do_sys_open+0x69/0x140
Oct 10 02:15:20 ovz6 kernel: [76546.471696] [] ?
compat_sys_open+0x1a/0x20
Oct 10 02:15:20 ovz6 kernel: [76546.506151] [] ?
ia32_sysret+0x0/0x5
Oct 10 02:15:20 ovz6 kernel: [76546.537500]
so what are you doing?
can we use our box now or wait?
would you tell us what to do rather than hang in the dark?
uptime – 19:13:06 up 14 min
this could be the best way to resolved all said issues including modprobe tun 😀
waiting for result and ticket to be answered.
wonder when will be available again…..=)
@John
There will be a downtime announcement posted here soon. During that time the changes will be made. The change (kern/OS choice) decision will based on information from testing that is currently being performed now.
thanks.
** Please read **
Both nodes in San Jose will be taken off-line for OS changes by 3:15PM CDT.
I expect this to take less than 30 minutes. Unless a unforseen issue occurs these nodes will be serving you by 3:45PM CDT at the latest.
hey, brian, would you post here as soon as it’s done?
Don’t know how people buy this, seems like a 4 days state of constant downtimes, if you can’t get any stability you should have beta tested before spelling such crap.
I would prefer a 1GB RAM OpenVZ VPS for $3.50/m as a backup server, or 512MB for $1.75/m…
It was tested. Very well. The problem with 2.6.32 vz patched kerns is that they crash under load. Not just normal testing load but real world customer load. They were flawless and performed *very* well during testing.
Just to update everyone — New identical nodes were provisioned this evening with CentOS 5.7 x64 and 2.6.18-274.3.1.el5.028stab094.3… 2.6.32el Cent,SL,RHEL patched OpenVZ kernels are simply broken. Period.
Everyone’s containers are currently migrating to the new nodes now. Have been for a couple hours now. They are live migrations unless they error then they are powered off before the migrate. I did not want to risk restoring containers manually.
This without a doubt will end the random stability issues.
Thanks for the update.
Thank you as well.
Sigh, as an Arch fan 2.6.18 kernel is a nightmare for sure. I just fixed the tty problem on Arch after update. Interestingly my VPS is at UptimeVPS but the solution was found on RAMHost forum and now I am discussing with eNetSouth. I’m going to trouble shoot another issue on Arch. Man I think OpenVZ should BAN all Arch users forever, so that we can simply use Xen or KVM, LOL.
I have to agree with you 100%.
My current issue is that I feel so let down with CentOS now and even OpenVZ to a extent. I know that OpenVZ kernels are built on RHEL kernels so it really does not matter much about your distro unless it is RHEL. What has me so mad is that they are taunting people to move to the 6.0 kernels. Calling them stable. I put my trust in the fact that it indeed was and no — not by a long shot.
I was reading where CentOS is basically in shambles and updates are taking forever. The big move now is to Scientific Linux. I tried that as well with this but it simply is a OpenVZ kernel issue that is carrying bugs from the 6.0 RHEL kernel. It’s all broken.
I’m not supposed to complain because it is open source. Right? Sure, it is open but people are depending on it functioning and it simply does not.
I embrace 2.6.18 with open arms. 🙂
any update bout tun?
Something went terribly wrong, I have 32GB of available RAM. 😮
i want to know update about tun too.
Anyone asking for Arch we just tell them to go with KVM and be done with it 😛
Brandon made a good call ditching .32 for the time being. At some point it’ll be stable but in the current state it’s a complete and utter mess and just scary to use.
Francisco
So I guess on 15th KVM would be my only option. 🙂
Well, there are OVZ templates that work on OVZ for anyone, but the latest/greatest glibc’s require a .37 kernel or something insane like that. While the OVZ devs have tried to backport certain features to make debian/etc work well, I don’t think the arch devs are able to really compile glibc different this time 🙁
Francisco
For Arch with a .18 host, I’m using glibc from here:
https://wiki.archlinux.org/index.php/VPS_Repo
But still the udev doesn’t work well. Also I am always worrying about breaking the system after upgrade since Arch repo is now on the 3.0.x kernel.
when you will provide IPv6
# free total used free shared buffers cached Mem: 32882224 249568 32632656 0 0 0 -/+ buffers/cache: 249568 32632656 Swap: 0 0 0I think they have bigger problems than tun/tap or IPv6 right now :S
If it were possible on OVZ then you might want to be a reseller and we can share the bill, e.g., $0.875/m for 4GB ram. I would like to pay some extra like $1/m for that. 🙂
I have more RAM than you n_n
total used free shared buffers cached Mem: 32882224 14436 32867788 0 0 0 -/+ buffers/cache: 14436 32867788 Swap: 0 0 0Not anymore! Ha!
Sad u_u xD
Btw, sorry but, any news on ticket #925807
This was because everyone was migrated from a 6.0 node that supports vSwap. Everyones config file had vSwap enabled and beancounters was maxed out 9999999999999 because of that. It is now set properly.
zzzzzzzzzzzzzzZZZZZZZZZ… When will they be stable and available again?
32GB of ram as well over here. Server rebooted again too :(.
[root@ovz7 ~]# uptime
11:50:39 up 12:04, 1 user, load average: 4.27, 3.45, 3.52
The last reboot was bringing the node back from adding the OpenVZ kernel at 11:49 PM CDT last night! Your VPS may have restarted due to the script that sets the properties right or from the migration. The nodes themselves have not.