| This is a review of my experience of 6 months using the WIRE-1 Reseller plan from WireNine.
Earlier this year I was looking to migrate one of my reseller plans away from a vendor with server and datacenter problems. I did research on WHT, and sent pre-sales questions to several hosts, including WireNine's CEO Ali. After review of his responses, their site and WHT comments, I signed up in April.
WireNine has a single reseller server that I know of, Lexus. Its performance and uptime is normal compared to other Linux/WHM/Cpanel accounts I have used over 5 years, and currently have. I use several paid and unpaid external monitoring services to watch domains hosted with WireNine, and although their uptime is not 100%, there is no exceedingly long downtime on Lexus server. With my monitors, I usually know within 15 minutes of HTTP problems occurring at Lexus server.
The forum or "community" at WireNine is slow to dead at times. Some hosts employing a user forum on their sites have active communities, but some never get highly utilized.
Point 1
Although there is a forum section "WireNine Network Status" labeled for "Announcements regarding service outages, server reboots, scheduled downtimes, upgrades etc.", most downtime & reboots are not recorded in this section (or any other that I see). After a reboot of Lexus that I noticed from Linux uptime reports, on May 9 I sent a PM to Ali asking, "Why wouldn't a reboot of a server (Lexus) be reported in the WireNine Network Status sub-forum?" and in this case he then posted a note in the forum. Other downtime, such as for RAM replacement on July 25, were neither announced in advance nor mentioned afterward.
My point 1 is that you cannot judge activity and events of the reseller server by looking in the WireNine community forum. Outages are rarely documented there. Don't scan their forum to make pre-sales decisions.
Point 2
WireNine maintains a status page for its servers. During an outage on August 20th, I noted not only that my sites on Lexus server was down, but my 2 personal nameservers provided by WireNine also were non-responsive. Surprisingly, the indicators for Lexus on the WireNine status page were all GREEN. I pointed these out in my ticket RZD-366300 alerting WireNine support to the Lexus outage.
Support responded: "The server was pingable, however web sites were down. That is the reason why the status page showed green for the server. The two name servers are on the server, them being up would not keep your web site online if the server went down. We are not running load balanced or clustered hosting.
We knew about the issue when it began, and we were investigating the issue since the time it began. The server was not down however, it just wouldn't load the web sites and then we had to reboot the server to bring the web sites back online. We will post an announcement on the forums once we know what initially happened. I pointed out that HTTP status green implies it can serve web pages. FTP status green implies it can transfer files. Your status display, only showing that ports can be pinged, is far less than what is implied. Their support's final response was, If the port is pingable, then it would show up as UP on the status page.
My point 2 is that at WireNine the green indicator for MySQL means no more about database readiness than the green indicator shown for POP3 indicates email functionality. The HTTP indicator will be green even if web pages cannot be served, as above. By the way, there never was a forum announcement after this incident. You also should remember that you are getting nameservers which are not independant as is recommended.
Point 3
In light of my uncertainty about how much WireNine knows of their server status, I sent a PM to Ali regarding their monitoring of server status, considering their Uptime Guarantee. My PM: What monitoring data is used to establish the uptime to determine if a web site is not available for more than 99.9% in any calendar month? Does that data measure only if a port can be pinged, or does the method verify that the web site is available (e.g. HTTP service is possible)?
Ali's response was, We use external 3rd party and internal monitors, they have nothing to with the status page available on our web site. We do not monitor the status page to find out if a server is down, we are aware if server is having any sorts of issues with our other monitoring services. Our external monitors log all downtimes, lag issues, etc. The 99.9% uptime guarantee has nothing to do with the status page on our web site, if you notice your web site is down you can request credit for it and we if determine that your web site was actually down at the time you mentioned and it was something we could have prevented then we will issue a credit on your account depending on the amount of downtime you suffered.
My point 3 is that other hosts that use announced, public monitors like Alertra, etc. have an advantage in providing public uptime stats and backing up their Service Level claims with verifiable data. WireNine is at a disadvantage showing, at most, the results of recent pings and not revealing 3rd party monitoring results.
Point 4
Ticket responses of "try again & let us know" while problems still active. Ticket IKN-396853 of 31 August was reporting a long-term email problem, when I wrote: Email sent to [domain] not reaching mailboxes nor forwarders; yesterday all incoming mail was delayed by about 5 hours. Can't tell how long delay is today; currently mail has not reached mailbox after 1 hour delay.
When I received the response, Please try again and let us know if you continue to face any issues., I replied: I think it is meaningless for me to be trying when the server status page accessed from WHM shows "red lights" for exim & imap, and server load over 9.
The next two posts from WireNine support were, "The upgrade has been completed, the load on the server has stabilized. Please let us know if you continue to face problems. You can ignore our web site's status page" and then "I am going to issue one month free credit on your account due to all the issues this month, as a token of our appreciation for your patience in this matter. We will continue to investigate this constant exim issue until it is resolved."
My final post was, Credit to me does not help my clients on this reseller account who cannot receive email. I have sent email to five of the Cpanel accounts on Lexus and none have reached their destination. The fact that this problem has not been mentioned in WireNine Network Status forum (to notify other resellers) is unfair and disturbing.
After this incident, there was no mention in forum and no credit to my account, but thankfully email service resumed. Why, oh why, was I being asked to "try again" when they have "this constant exim issue"?
Point 5
Do not uniformly maintain reseller servers with updates.
The WireNine web site states they use the latest stable build of PHP. In the forum it was pointed out as late as August that the PHP upgrade of May had not been applied. Also, after Lexus server was updated, it took another requestor to get further servers updated.
Point 6
Major service changes take place without warning.
Even though there was a strong commitment to announce Cpanel 11 upgrade before it happened, in reality Cpanel 11 was fielded without warning.
My point 6 is that WireNine does not understand that for resellers, it is important that advance notice/planning is done so changes can be communicated to the reseller's clients. This is a difference from a simple shared (individual) hosting plan. Major service changes cause resellers to have a flood of phone calls from clients when skins change or scripts break.
In addition, users of Spam Assassin found that email subject re-writing (e.g. inserting a SA score) suddenly stopped. In ticket FCU-300143 I pointed out that the Cpanel pages no longer offered setting re-write. Support first said it was due to Cpanel skins, then claimed "that configuration is not available in the new version of SA". After I sent them a screenshot from another host with a Cpanel 11 X3 skin and current version SA offering subject re-write, the truth came out: "Spam filtering is now done at SMTP time as an ACL, This prohibits subject rewrites from working". Again, WireNine had changed the functionality of a tool, Spam Assassin, without notifying hosting clients.
In the last weeks, I have found their support center frustrating, as there is a tendency to explain away issues rather than attack them. As a ticket develops, Support's explanations for problems tend to change. Additionally, I was surprised when support asked ME in a ticket, "Where are the MX records pointing for the domain?", since it seemed someone in a support position should be able to determine that while working a ticket.
My final opinion is that WireNine is NOT the place for reseller hosting, unless you want surprises and excuses. My doubts about what would happen next(?!) led me to leave their reseller hosting plan. Perhaps the situation would improve with WireNine taking a more active and honest role in communicating with their clients.
|
Shared Hosting
Shared Hosting
Add to Favourites 
Print this Article