Article Last Updated: 16 Nov 2018
$1 per month / $12 per year list price / 1 CPU Core /
128256 MB RAM / 1015 GB Disk
In December 2016 we’ve moved to a more powerful ChicagoVPS OpenVZ Professional plan (re-branded to Enterprise in 2018):
$6.50 per month / $78 per year list price / 4 CPU Cores / 1 GB RAM / 30 GB Disk
Performance and stability of the new server is much higher (zero downtime / (un)scheduled reboots in the past ~18 months!).
We’ve just thought this should be worth knowing, as it seems that their basic plans are completely over-crowded, with stability and downtime issues, not present in their more expensive offers.
Recently (in the past 2 weeks) our Enterprise server (ex Professional) started showing some hick-ups, without any significant load increase whatsoever. Support was usually blunt about this, simply “rebooting” our server to fix the “non-existent” issue. Hopefully this is a temporary glitch in their node, but it can also become a regular thing. Could be related to recent MySQL server update, as some experienced problems. Sigh :(
Epilogue: We had to rebuild our Enterprise server from scratch (yikes!), but at least support team provided us a free full backup.
According to CVPS tech support team:
It happens generally because of high resource usage and high load on the node.
(translation: overselling / overcrowded service)
We continuously monitor average load to maintain host node load to avoid such issues.
Obviously, not as effective or continuous effort to provide great QoS :(
ChicagoVPS offers some competitively priced VPS and Dedicated Servers hosting plans (along standard variants such as classic Shared hosting, OpenVZ/KVM VPS, Linux/Windows OS platforms etc.). After half a year full year of using their services we can put a real, unbiased review of their cheapest VPS server.
We use the cheapest $1/month VPS plan with following official specifications:
128 RAM | 10 GB HDD+SSD | INTEL Xeon CPU E3-1240 V2 @ 3.40GHz – 1.00 CPU core
We were surprised that we actually received slightly stronger setup for the same price:
256 RAM | 15 GB HDD+SSD | INTEL Xeon CPU E3-1240 V2 @ 3.40GHz – 1.00 CPU core
(double amount of RAM, no swap, 50% increased disk space)
UnixBench 5.1.3 @ Ubuntu 14.04 score = 2432
115 ~ 193 MB/s
CPU is very decent, for low to medium traffic with website that serves mostly static content this will be more than enough. For some intense MySQL operations: forget about it (unless you expect < 1000 visitors per 24h).
According to our benchmarks, disk system is either some kind of a conventional RAID setup, or SSD-cached one. The results greatly vary from test to test, probably highly dependent on the NODE overall load, but we got between 115 MB/s to 193 MB/s with sb-io-test, not high-end, but very decent given the price.
There are other benchmarks we performed, but they highly depend on server configuration, type of scripts/applications and traffic, so we are not gonna discuss them.
SERVICE UP TIME
Overall, it is up to the specs and we are very pleased by this. There were 2 major downtime periods that we’ve noticed during the first 6 months of the service use, lasting for at least couple of hours, but unfortunately, nothing was posted on their website, Facebook or Twitter. Assuming it was not a planned downtime.
Update (September 16, 2016)
After 1+ year of service, we can say that there are indeed some service “soft” downtimes. By “soft” we mean that some users on the same NODE ( = yours ) occasionally may abuse ToS and CPU usage over time. This will consequently affect all VPS servers on the same NODE, manifested through long server response times, frequent timeouts, SSH and SFTP authentication and access failures, and more. The VPS will be (technically) still fully operational, however, due to high CPU throttling caused by other abusers, it will pretty much affect everyone.
This is probably a limitation of OpenVZ architecture they are using, where clients are not 100% truly isolated from each other in terms of resource sharing and usage, but it is also a clear signal of possible NODE overselling issue.
The best advice you can get regarding this issue is to immediately open a high-priority support ticket and report abnormal CPU usage, high downtime and poor performance. Staff will hopefully promptly respond and take action.
Update (February 05, 2017)
Near the end of year 2016. we’ve received an email from ChicagoVPS that something big is going to happen, and that some improvements are on their way. At first, we’ve just ignored it and thought it was another end-of-the-year marketing campaign.
However, soon we’ve noticed that server stability was greatly improved. There were no more “soft downtime” issues, and at first we’ve thought that this was probably because of the holiday season and all. The ChicagoVPS staff increased all nodes capacities to avoid problems during the break. Now, we are long past the holiday season time and our server stability is still better than it was before. Take a look at our most recent CPU/RAM usage chart (and compare it to above soft downtime version):
The CPU spikes are rare, sometimes they are well below 1.0 value. All looks normal.
Coincidentally, with this obvious upgrade, we’ve noticed 2 new symlink files in our server’s root dir: aquota.group and aquota.user – take a look:
We cannot access those files (system reports that they do not exist!), but they probably have something with above mentioned NODE overload issues in the past. Some users were abusing the ToS agreements and causing overall server overloading, affecting all other VPS containers. With this upgrade, the resource distribution seems to be much better controlled and fair.
Now, this is all pure speculation from our side – there is no official confirmation regarding any of this, and we will continue to monitor our server performance and uptime / availability.
Update (February 24, 2017)
And CPU spikes / RAM dropouts are back… Oh, well.
The problem now is that we’ve moved all our high-traffic websites to a more powerful plan (where this kind of issues no longer happens), and it should definitely not happen in this system anymore, as this low-end server now hosts only handful of extremely low-traffic websites.
This indicates that someone is most probably abusing the node resources again, and we have to open a support ticket to fix it.
FTP UPLOAD RELIABILITY & SPEED
Since we are in Europe, and server is located in New York, there is a noticeable lag when we upload files via SFTP. We have much better experience with VPS and Shared hosting servers located in Europe in this regard. Sometimes there are strange timeouts reported inside WinSCP, so if you rely on heavy FTP work, you might first check and make sure that your connection is up to the level you expect.
Please, don’t get us wrong here. FTP is reliable, say, 99% of the time. It works fine, but sometimes, when transferring and quickly editing small files, they might not upload correctly and in a timely manner, so you will have to restart the transfer process inside WinSCP (or other FTP client you use).
VIRTUAL SSH RECOVERY CONSOLE
Virtual Console is a bit clumsy, it requires JAVA installed and running (which is a major issue, since Chrome removed support for NPAPI plugins entirely). If you lock yourself out from SSH access, you will have to downgrade to an older Chrome (portable edition or ran inside virtual box or sand box), use Firefox (scheduled NPAPI support removal in 2016), or even Internet Explorer / Spartan. They will probably resolve it in upcoming months.
update : Dashboard was updated to include HTML 5 Console, however it does not really work, always returning connection aborted – missing connection key error message. If you ever get locked-out from the console, opening support ticket is your only way out of trouble (assuming they can help you in this case – not verified)!
update : At least you can use it to generate temporary password and access your server in case you lock yourself out of regular ssh access. But, instead of web interface, you must use your own SSH client (e.g. PuTTY on Windows). Sessions are limited to 1 hour (60 minutes).
USER / CLIENT DASHBOARD
Very simple, provides some basic stats (memory/cpu/network) via RRDtool. However, what most users might not realize at first is that there is a Control Panel which you must access via drop-down selector
Product Details > View Product > Management Actions > Control Panel > Manage > Statistics
from where you can access entire statistics history (with granularity from 1 hour to 1 year). You cannot bookmark Control Panel URL and access it directly.
Overall, it is simple, well-organized, but some key areas could be more visible. There is nothing advanced in there, besides setting up your instance/server name, shutdown, reboot, rebuild…
CAREFUL: there is NO protection mechanism if you accidentally click on Rebuild button! Oops!
(update: fixed as of Sep/20/2016)
Update (January 30, 2016)
ChicagoVPS put a nice face-lift of their website, including User Dashboard. They’ve moved to bootstrap template to refresh older design a bit, but the functionality is the same. However, now some areas such as Control Panel are more easily accessible and visible (but Control Panel itself is still very old in appearance, it was not updated).
There is one important change: Operating System and Location of the VPS instance are now moved into Configurable Options tab section. However, we could not change any of them. Naturally, you need to rebuild entire VPS server to change operating system, and we are not sure about location, but we assume you need to contact support if you wish to achieve that. So, in effect, this is probably useless for anyone.
Update (September 20, 2016)
ChicagoVPS put another, smaller, facelift again in User Dashboard / Control Panel. New large icons with sections were added, emphasizing most important commands, and moving statistics to a separate page (directly accessible via bookmark, for example).
The most important change today is introduction of separate pages / sections. It is no longer possible to accidentally hit REBUILD button and wipe entire VPS server out!
There is no longer access to old / legacy Control Panel, which had one option modern panel does not: RRDTool stats for up to 12 months worth of data! You could plot the chart of your server CPU Load, Memory and Network usage and get a pretty good picture of what was going on. Now, now you are limited to a fixed chart with last 24 hours only :(
Additionally, they have incorporated CloudFlare CDN service, namely, DDoS protection, so now users must wait up to 5 seconds before accessing the login page (if CF session is expired).
Finally, latest change introduced a new COMODO CA Limited SSL certificate, which is unfortunately NOT compatible with older Chrome browser versions and Windows XP operating systems. Their previous certificate until last week was working fine, though.
Update (April 23, 2018)
ChicagoVPS put another, smaller, facelift again in their website, essentially entire User Dashboard / VPS Management remained unchanged, but the header, footer and overall menus across the entire website were updated with new theme.
CUSTOMER SUPPORT: SPEED & QUALITY
Well, we will be honest here: it is NOT lightning fast. Sometimes, we had to wait 1-2 days for a reply, which is SLOW by all definition of support speed. However, we did not had any major issues with our instance that required immediate attention from support staff, so that is a big plus. If you heavily rely on ~ instant support replies, seek something else.
EMAIL SERVER ISSUE (Blocked SMTP Port 25)
Update (March 11, 2016)
As Murphy’s Law goes, we’ve finally got hit by a major issue. And after a whole week (problem surfaced on March 5th, still unresolved as of today), apparently entire VPS NODE, which includes our own VPS server IP Address, was marked for spamming issues, thus our VPS server’s port 25 (outgoing email traffic) was blocked with inability to send any emails at all. Email traffic ceased both for users [registration / validation purposes], and admins internal system notifications [available updates / comments awaiting approval etc.].
At this point in time, only one thing is left to be said: we consider ourselves really lucky for not using ChicagoVPS for any “serious” business application that relies on emails and user registration, otherwise, we would be at real loss with no email traffic for a whole week. During our conversation with tech support, we provided whole bunch of proofs that there is 0 spamming activity on our end, including very low email traffic log, and no listing of our server IP on many popular spam databases in the world. In the end, who in their right mind purchases a server in advance for 3 years, just to risk account closure for spamming?
At this point in time we cannot describe the entire issue in details (it will be posted in a separate entry once we gather all the facts), but this has clearly lowered our service satisfaction level and will affect our final review verdict, with additional warning issued.
Update (March 11, 2016 in late afternoon)
The issue with blocked port 25 was finally resolved! We are not quite sure about what was the cause of all the trouble, since nothing in our server environment changed (not even IP Address, as in some other client cases did), so very little details to build-up scientific guess work & theories. Still, 7 days without email traffic could be a disaster for business.
We can recommend ChicagoVPS to developers and small website projects (remember: we have ONLY reviewed their basic VPS plan here, not the more expensive VPS and Dedicated Servers plans). Overall, we are very happy with the setup, although, there is certainly room for improvements.
We cannot 100% safely recommend ChicagoVPS for business and time-sensitive services, due to the fact that in some complex cases they need to work with their upstream provider, which means long delays in problem resolutions (please read our March 11th update above).
- SFPT file transfer could be faster, for some reason it is not 100% reliable in our experience. Occasional problem with timeouts and delays will occur (this may be influenced by your physical location and distance)
- Support replies must be faster (within 1-2 hours at most) (update: this was greatly improved)
- Problem resolution can be extremely slow – in complex cases it depends on ChicagoVPS’s upstream provider, which may take days or weeks to be resolved, if ever
- Occasional NODE overloads and VPS ToS abuse by some clients will affect your server and service(s) as well – contact support in High-Priority mode for their immediate attention
- Extremely cheap/affordable!
- Simple User Dashboard
- Reasonable service up-time (use higher / more expensive plans if you need better stability)
- Great performance of a budget entry-level VPS