I've been with ServerWizards for the past year. Prior to contracting their services I emailed them regarding some of the policies and issues I was concerned with and were not address on their site
Response time 5-10 minutes
Resolution time 2 hours, "usually much faster" but allowed for extra time in more complex issues
Update php, mysql etc "as soon as there is a crucial update"
Security scans and server health scans every weekend, and minor updates applied
Compensation if there is downtime cause by their staff
Fix downed services immediately after monitoring software picks it up
I mention this because during my stay with them, I don't feel most of them have been systematically addressed.
Most recently (on the 31st of May, 4A22461) I asked them to switch from Apache to nginx. The reason of this being that my server kept crashing specially during peak times. After observing the crashes first hand, monitoring the resources etc, I came to conclude it had to do with ram, stopping apache would abort an infinite loop in which ram would be deapleated, swap would b eused and the load would skyrocket for hours until the server eventually crashed (hopefully) and ServerWizards would pick it up and do something about it. Unfortunetaly merely restarting the server is not enough, since apache will do the same loop unless stopped and started few minutes later.
Anyways, I opened the ticked on the 31st, now almost 3 week later remains unresolved. I asked the to do backups prior to installing nginx on the 1st of June. It took 3 days for the backups (local and remote) to be done, kind of pointless when 3 weeks past by those backups... if something goes south.
On the 4th they wrote that they were starting on a compilation.
On the 5th they said the compilation should be completed today.
On the 6th. The server spent 8 hours down (httpd down). They promised it would be done on that day
On the 7th. There were some issues with nginx displaying php (according to them. They had started configuring vhosts).
On the 7th. I informed them I was confused as I couldn't find anything remoretedly relating nginx using find, locate and whereis although they had claimed to have performed certain benchmarks...
On the 8th. Information relating to nginx would be provided
On the 10th. Apologies, asking for more time (24h) because they were going to "recompile the majority of the software on my server for better performance". during those 2 days, my server httpd was down, the load made ssh useless, and I had to fix it myself...
On the 13th. More apologies for the delay [note: no communication in the mean time], they had serious issues in their datacenter and would start work on my server now
On the 15th. Ready to install nginx, asking whether I want more modules
On the 16th. Sent PM to KindAdmin here hoping to avoid this thread (no reply yet)
On the 16th. Further apologies, and "we will rush to get this completed for you"
On the 17th. Compilation finally installed, starting on configuring nginx.conf
On the 18th. Asked how much longer would ti take
On the 18th. Opened a ticket (4A3ADD40) with High priority (original ticket was at Low as I wasn't expecting such slow responses) complaining that my server had spent 6 hours down (httpd again) and reminding them of what I was promised when I had contracted them
On the 19th (12 hoours later). Indicate they are in the final stages of configuring nginx.conf (no modifications have been done to the file as of 20/6/09)
On the 20th. They finally have it “ready”, ask me when would be a good time for me so I can test the website (running vbulletin). I replied within 25mins indicating they could, and provided them and IM method for faster communication if necessary (they haven’t used it).
On the 20th. 3 hours later, they indicate nginx has been implemented. Logins do not work. Some more back and forth communication
On the 21st. Ask them to store nginx logs using symlink on another drive, and rotate logs. Can’t verify whether the rotating logs has been done. I asked for the commands to start and stop nginx, what they originally gave me would not stop neither nginx or fastcgi…
In the 22nd. Try another unsuccessful test… php not rendering properly
On the 24th. Opened another ticket (4A427131) because the server had been down for a while, and kept almost crushing as I kept stopping apache trying to login to the forum Admin control panel and disable the forums… I expected it to be a rough night… unfortunately I only had dial-up. The said they were working on it -.-
On the 25th. Wrote them very upset after waking up with the server crashed (peak time is sleep time where I live due to time zones)
On the 25th. Asked for the progress on the ticket (4A3ADD40). No communication done during those 3 days…
On the 25th. They claimed to be ready for the switch and required me to check the website. I didn’t reply until the 28th as I had no proper internet to do the testing with.
On the 27th. I gave them the information for a installation of ngaios I had done months ago so they could add their email and monitor my server, whatever the hell they used was utterly useless when httpd crashed but the server pinged… I may have not been clear, but I believe they disabled my email form getting notified in the process.
On the 28th. Informed them that I could do the testing, they said they would run the tests, I said okay, next thing I know the guy ask me whether I want ngaios and another software configured too (vhost). I said yes, but first things first, make sure php works properly, they said ok…
On the 1st of July… nothing has been done yet. Not only that, but … the server has enjoyed over 5 hours down after it was rebooted (don’t know by whom or why).
In the meantime this is the story of my website for the past month. Numbers indicate the number of hours down.
May was very stable, April wasn't as bad as June is being shown to be but far from great.
Anyways this is the latest issue. In addition it took 12 days to update php, apache and mysql. Their monitoring system never picked up when httpd stopped working, I think they only monitored ping... their response times overall have been very disappointing, to be frank I think my response times average far better than they do, most of their replays were at about 7-9 am CDT and then I could wait another day to get addressed again.
In addition to this issue, overall I did not feel they were proactive regarding security, They were rarely if ever on the last login, I was not informed of any updates if any during my stay. I just felt out of the loop, and they outside my server, not even looking (out of sight out of mind)
On a positive note they have been great regarding helping me add new dns for new websites, good at loading up backups and doing symlinks. However this does not compensate for the hours and hours my website has been down, and the unfortunate lost of trust in their work this has caused. I'm hoping this review will help them shape things up and improve.
I feel like I’ve been played around like a worthless toy…I refrained from submitting this post on the 21st as I had planned) hence the outdated picture)… but it is just ridiculous now.
Hopefully they wake up soon.
PS sorry to SW on some of my latest tickets as my tone got a bit higher than the professional etiquette demands... Hopefully seeing it from my perspective they will understand the frustration of seeing your server falling non stop during that time, and the cost of money and loyal users that entails
Quick update, according to serverwizards, they haven't touched nagios. I started the service and apparently it was off (I wonder how... as it starts with booting too) apologies for the accusation, it just seemed in line with everything that has been happening....
The server was not doing too well even before I asked them to even try to install Nginx. If you compare the dates I have given with the graphs, you'll see that it was dropping for hours no end well before they did anything much beyond backups. Nginx was supposed to be my fix. As the problem seemed to be regarding how much RAM apache consumes...
I'll attach you a copy of what April looks like (disregard the 30th). May had short fall here and there throughout the whole month, but is not visible in the month view of the graph
I'm in the process of looking for a management company capable of taking care of their clients, got one very good one, but looking at alternatives too.
Definitely give another server management firm a try. Also, why don't you ask your current server provider to manage your server? Server providers tend to put more efforts in pleasing their customers especially when their whole operation's reputation is on the line.