Web Hosting Talk : Web Hosting Main Forums : Web Hosting : PrecisionEffect Review

[mofahead]

I thought I'd share my experience with PrecisionEffect over the last year or so...

After some consultation on WHT I chose PE as a reliable, fast, shared host in Jan 2009. A bit more expensive than my previous host but I was hoping that the speed and quality of support would compensate.

I have to say that at first, PE delivered with fast page loads and helpful service.There were a few kinks here and there, but they were ironed out promptly. For the the one or two short server outages, we were notified in advance.

The problems started in January this year. Now I wasn't even aware of the management change, as has been discussed here in other posts, but it appears that the company had been acquired and the changeover had taken place at this point.

In any case, we were hit with a database failure in mid-Jan that brought our site to halt overnight. It turns out that this was the result of periodic server maintenance, according to the PE staff. However, we were not informed in advance, and only after we sent in our distress call was our site restored. In response to this debacle, PE decided to upgrade us and everyone else on that server to a new server with failover protection, supposedly equivalent to their $50/month business package,

Last week many of PE's customers, along with my company, discovered how effective the new failover system is: that is, not at all. Not only was our website went down unexpectedly and staid down for over 48 hours. Email, ftp, and the control panel were all down as well. Support was both woefully slow in informing us what was happening, and inaccurate in it's forecasts for renewal of service.

Our has since been restored and everything appears to be back to normal. The official explanation lays the blame on the datacenter. We were offered three months of free service (for which we have to submit a ticket to redeem), and there is also a claim of another server upgrade.

Unfortunately, despite these promises, my company has lost faith in Precision Effect. We'll be moving as soon as possible to a provider who can deliver more reliable service.

[IGXHost]

I am sorry to hear about your poor experience with PrecisionEffect. They seem to have maintained a pretty positive reputation here on WHT. I've taken a look at their failover plan which is rather pricey for shared hosting but you would definitely expect what they're advertising. However, this review has got me thinking otherwise.

Have they provided you any reasons and explanation of why the downtime and other services were down as well?

[mofahead]

Originally Posted by igxhost

Have they provided you any reasons and explanation of why the downtime and other services were down as well?

Yeah. This:

"I would like to thank everyone for their continued patience and loyalty during this catastrophic downtime. We felt the move to a redundant infrastructure w/ iSCSI SAN was the best move for our clients. However, this all came tumbling down when the datacenter we are in reported massive SAN input/output issues. Hewlett-Packard Storage Works believe the issue was isolated to the storage nodes we were located on in the SAN." -PrecisionEffect CTO Ronny Fang

Since I don't really understand the jargon, this excuse or that isn't really the issue as much as whether the company can provide reliable service. And though it's possible that the failures we've experienced (you can read some other reports on this website) are the result of bad luck, rather than bad design, I just don't feel confident going forward with this provider.

Hopefully some of those people who are sticking with PrecisionEffect will share their experiences in the future.

[Quartz]

There have been several threads popping up in the last couple of weeks regarding the recent changes withing PE's management and their services provided.

Another one seems to have bit the dust here. Obviously transitions can come with its bumps but this is unfortunate for all parties concerned.

What domain(s) did/o you host with PE as to confirm this review? You can click report and submit privately if you wish.

I can only imagine more threads coming regarding their recent string of events. Hopefully they can move forward and bring back the past of PE that many know.

[hD-Scott]

It's sad to hear that another one of my DirectAdmin brethren have taken a turn for the worse after a management change. First VectorLevel then PrecissionEffect. Hopefully us small guys can stick it out!

[Ryan - HostATree]

I use to use them way back in the day for a nameserver vps. Always good for me until they couldnt accommodate the custom needs I required. But this was waaay back.

[ItsJustHosting]

Originally Posted by powerMonster-Scott

It's sad to hear that another one of my DirectAdmin brethren have taken a turn for the worse after a management change. First VectorLevel then PrecissionEffect. Hopefully us small guys can stick it out!

Indeed it is, they've built an exceptional reputation for themselves, it's sad to see it all jeopardized.

OP: did they give you a reason for why you were down 48 hours?

[PE-Phil]

I would like to go ahead and post our message in this thread as it seems a lot of people have questions concerning what happened and what has been done about it.

This is a message we sent to all of our clients. I believe this should clear up many of the questions you have in this thread. We are so sorry that this has happened, and we're thankful that all data was able to be fully recovered.

What a week.. I cannot begin to express my sorrow at what transpired this past week.

At 12PM Wednesday afternoon we began to experience high load issues on our Cloud Infrastructure which hosted 2 of our servers, AlphaCentauri and UrsaMinor. By 1PM sites were going offline due to the extreme iowait times. Iowait is basically a bottleneck in data transmission from the storage device, the SAN.

In essence, data was being delivered so slowly to the server that no web pages were capable of being served. We diagnosed the cause of this problem to lie with the clustered 40TB iSCSI SAN provided by our datacenter (using Hewlett-Packard StorageWorks). The datacenter spent precious hours troubleshooting it to no avail. By midnight that night, they determined it was an issue with our cloud, that we were sending mass amounts of data to their system.

We knew this was not true... we host regular clients with dyanmic yet normal content. By this time the downtime was already unacceptable. They claimed that we were the only client having issues. I spoke with the CEO of our datacenter and convinced him to move us to another cloud.

By morning, the migration was complete yet the problem persisted. We were forced to immediately switch all of our customers to traditional boxes running 15K RPM SA-SCSI Drives. The sites were restored from backups that were a day (by that time 2 days) old. However our sync of the latest, fresh data from the SAN was being transferred at crawling speed due to the iowait issues. By this time the datacenter techs informed us other customers were having similar issues. Friday morning Hewlett-Packard technicians were in the datacenter troubleshooting and diagnosing their SAN.

As of this moment we still have no had an explanation from our DC what the issue was, or if it is fixed. The last we heard was that there may have been faulty drives in part of the SAN - coincidentally the part of the SAN that we were stored on. This has been an arduous journey. Many members of our staff stayed after and even stayed for double shifts, without even asking for overtime, during the past few days. It makes me happy to know that we have staff who care so much. Phil and I have not left the office since Wednesday night. We are completely invested in this company, and it breaks our heart to see so many of our clients websites go down.

We know how important your websites are, and how stressful this entire situation has been. I know that we have lost the trust of many of our clients, but we will continue to be here and build back your trust, one day at a time. For those of you who inevitably decide to seek an alternate provider, please send in a ticket for a full refund of February. You should not need to pay for a month when service was not met. We accept full responsibility for what happened, and we'll also hold your account for as long as you need to move data off of it. For those clients who decide to stay with us, we thank you and appreciate you from the bottom of our hearts. We promise we will not let you down again, ever.

The clients who have decided to stay with us, please don't forget to send in a ticket to Billing to have the hosting charges for the next 3 months waived.

If you have any questions, concerns, or comments, please feel free to email us, call us, or send in a support ticket.

Best Regards,
Ronny Fang, CTO
615-500-0317
PrecisionEffect.com

on Behalf of Our Entire Staff
Philip, Alby, Angela, Aaron, Kris, and Roshan.

[Disrelation]

Wow, this is very very unfortunate and as mentioned above, they have seemed to keep a pretty good image here on WHT but this does having me thinking twice.

Well, I didn't go through the whole letter but it seems they're trying to make it up to customers.

[routingcast]

According to that message, you have no idea what the problem was! This could very well happen again!

Originally Posted by PE-Phil

As of this moment we still have no had an explanation from our DC what the issue was, or if it is fixed. The last we heard was that there may have been faulty drives in part of the SAN - coincidentally the part of the SAN that we were stored on.