By: GZ on Venerdì 29 Luglio 2011 00:18
Non lo so, mi sono fatto un poco una cultura negli ultimi 4 gg su come funzionano 'sti server o perlomeno come dovrebbero funzionare.
------------------------------------------------------------------------------
Questo è solo uno dei cinque ticket di assistenza online in chat che ho riempito dal 24 luglio
-------------------------------------------------------------------------------
Ticket Contents:
User: XXXXXXXXXX- 2011-Jul-24 11:42 (GMT-0500) [Update 1]
I requested yesterday a reboot and also a diagnostic check of my dedicated server for my website www.cobraf.com because it was extremely slow to respond to any browser and also when I was connecting by Terminalserver. Basically the website was not working, it showed only pieces of the main home page and of the frame and did not open any page
I got this morning the following message:
"...The diagnostics have completed and it found a correctable ECC error for DIMM 3. I have replaced the RAM module for DIMM 3 and this server is now back online responding to ping, http and RDP. Please let us know if there is anything else that we may do for you.
David L.
Server Build Technician
SoftLayer.."
But the website is still very slow and I had to stop Mysequel in order just to connect with TS from Italy
Is there any other hardware check you can perform or any other monitoring or diagnostic of my server ? At the minimum can you do another hardware check ?
Employee Response - 2011-Jul-24 12:10 (GMT-0500) [Update 2]
Hello,
Let me check your query. I shall update you soon. Please hold on.
Best Regards,
Rey R
SoftLayer Support
Employee Response - 2011-Jul-24 12:19 (GMT-0500) [Update 3]
Hello,
I am unable to login into the server. While I am trying to login into server it is showing the message "Terminal server has exceeded the maximum number of allowed connections". Please refer from the attached snapshot. To proceed further, please allow us a free RDP session for logging into the server.
Awaiting your reply.
Best Regards,
Rey R
SoftLayer Suppor
File(s) attached to ticket RDP.png
File Attachment:
RDP.png
User: XXXXXXXXXXX - 2011-Jul-24 14:09 (GMT-0500) [Update 5]
I Think you can now connect through Terminalserver
Employee Response - 2011-Jul-24 14:29 (GMT-0500) [Update 6]
Hello,
I could see that the website 'www.cobraf.com' is loading absolutely fine now. A spot check at http://alertra.com revealed that the domain is loading perfectly fine from all over the globe.
However, you can do a passmark test on the server for checking the performance of the hardware components. The Passmark utility is a software solution for stress testing a server's hardware, as well as running performance tests, drive checks and monitoring changes to sensitive files in the OS.
Burnin will stress test the hardware for a pre-determined amount of time, forcing latent hardware errors to the surface. It's important to make full backups of all needed customer data prior to running this test, as it will force faulty hardware to fail completely. You can initiate a passmark test through portal:
-------
Portal>> Hardware >> Configuration >> Initiate a Passmark hardware test
-------
Please refer these links for getting further information:
-------------
http://knowledgelayer.softlayer.com/questions/266/What+is+the+Passmark+Utility%3F
-------------
Feel free to contact us, if you need any further assistance.
Best Regards,
Jayce M
SoftLayer Support
User: XXXXXXXXXX- 2011-Jul-24 15:47 (GMT-0500) [Update 7]
Hi, We were conncted by TerminalServer for a while so that's why you could not get in a couple of hours ago.
We would like to ask
i) another check of the Hardware and also
ii) to discuss with you an upgrade of it. Please understand that is a very serious situation for us, our forum is basically not working since Saturday. Here are some more data for you
We see the CPU of our dedicated server running at 100% most of the time now and therefore our website is stuck and very very slow. We need you to check the Hardware again though, because we performed several tests today and they indicated that the machine might have problems beside the one you found yesterday (see previous ticket).
First we notice that when MySql is stopped CPU usage drops immediately and the website is fine, but obvously out forum works on a MYsql database. So we reinstalled MySql on the server, then we also copied it into our pc here in Italy and launched the same queries that it performs on the website on Softlayer. We see that on our local machine here the CPU DOES NOT get even close the CPU 100% usage like on your machine at Softlayer.
Therefore we need to check again the dedicated server HW
Employee Response - 2011-Jul-24 16:01 (GMT-0500) [Update 8]
Hi,
I shall escalate this ticket to our senior technicians. Please standby, you will be updated shortly.
Best Regards,
Jayce M
SoftLayer Support
Employee Response - 2011-Jul-24 16:51 (GMT-0500) [Update 9]
Hello,
I am currently further investigating this issue. Please stand by for further updates.
Thank you for choosing SoftLayer
Ira M.
Customer Support Technician
http://www.SoftLayer.co
Employee Response - 2011-Jul-24 17:41 (GMT-0500) [Update 10]
Hi,
I have noticed no hardware error messages in the server's System logs but another hardware check can be scheduled for your server. We currently have the following scheduled maintenance windows available for non-emergency hardware maintenance:
-- Any day of the week, 9:00am - 12:00pm CDT (GMT-5)
-- Any day of the week, 5:00pm - 8:00pm CDT (GMT-5)
-- Any day of the week, 1:00am - 4:00am CDT (GMT-5)
Please let us know which maintenance window works best for you. (If necessary, we can schedule a custom maintenance window for you, but please be aware that we will not always be able to accommodate every specific request.)
Concerning MySQL using most of the CPU performance, the problem may be caused by inefficient queries. Can you explain the layout of the database and show us your sql queries and/or stored procedures. Also, can you update the ticket with the MySQL credentials.
Thank you for choosing SoftLayer,
Ira M.
Customer Support Technician
http://www.SoftLayer.co
User: XXXXXXXXXX - 2011-Jul-27 17:34 (GMT-0500) [Update 11]
My website is not even visibile since a few hours today, since Saturday my machine has been either offline completely like now or visible but stuck, with CPU at 100% all the time
I already filled 5 tickets since Saturday asking a HD diagnostic, 2 reboots, a full chassis Swap and an I Cisco Guard
What can I do ?
Employee Response - 2011-Jul-27 17:57 (GMT-0500) [Update 12]
Please standby as I reboot this system via the Orbit ticket open then I will update this ticket and send to support.
Thank you,
Ricardo C
Server Build Technician
www.softlayer.com
Employee Response - 2011-Jul-27 18:12 (GMT-0500) [Update 13]
The system is now back online.
It was at a blue screen showing RAM errors.
I will need to replace all the RAM.
May I proceed or would you like to schedule this down time?
Thank you,
Ricardo C
Server Build Technician
www.softlayer.com
User: XXXXXXXXXX - 2011-Jul-28 05:14 (GMT-0500) [Update 14]
Hi, I requested some 15 hours ago a Ip Cisco Guard because my website www.cobraf.com as basically shut and I could not even reach it with Terminal Server and we had detected since Saturday a suspicious activity from some US IPs(We are in Italy) requesting dozens of pages simultaneously all the time
The website is up now, but it gets disconnected every 10 minutes, we are monitoring and we see as if you unplug the network cable avery 10 minutes. Is this what the Ip Cisco Guard is supposed to be doing ? Is it doing this then because it is detecting some attacks or just as a default procedure regardless ? Can you please not take OFF the Ip Cisco Guard so that we can see if the server stands up alone ?
Employee Response - 2011-Jul-28 05:28 (GMT-0500) [Update 15]
I am looking into this. Please standby for updates.
Regards,
William B.
Server Build Technician
SoftLayer
Employee Response - 2011-Jul-28 05:47 (GMT-0500) [Update 16]
The guard is still active. It is possible that the cause of the network outages is the maintenance/poking you are doing with the server is causing the guard to see your IP as hostile.
Please let me know if you continue to see this behavior.
Regards,
William B.
Server Build Technician
SoftLayer
Employee Response - 2011-Jul-28 05:52 (GMT-0500) [Update 17]
Hello and thank you for the phone call. I also saw the system offline and am looking further into the issue. Thank again.
Barrett C.
SoftLayer CSA
User: XXXXXX - 2011-Jul-28 06:02 (GMT-0500) [Update 18]
I attached an image of the Tracert www.cobraf.com that shows it does not reach the server and then 2 minutes after it does and so on. Is it a problem of connection through this node coming from Italy ? As you can see from the tracert it stops ar the last node
can you please check it ?
File Attachments:
tracert.GIF
tracertcobrac.com
Employee Response - 2011-Jul-28 07:26 (GMT-0500) [Update 19]
Hello and thank you. We have removed the Guard protection and am monitoring the server. We will have an update as quickly as possible.
Thank you again.
Barrett C.
SoftLayer CSA
Employee Response - 2011-Jul-28 08:07 (GMT-0500) [Update 20]
I have been watching the server and it appears to be stable.
--- 67.15.18.55 ping statistics ---
900 packets transmitted, 900 packets received, 0% packet loss
round-trip min/avg/max/stddev = 4.610/6.866/39.282/2.433 ms
Please advise if you see otherwise.
Thank you.
Barrett C.
SoftLayer CSA
User: - 2011-Jul-28 15:13 (GMT-0500) [Update 21]
Thanks
the server has been running OK again in the last 5 hours, so I am glad of this onthe other hand I do not understand what happened in the last 3 days exactly, even this morning ?
What do you think caused the server to be disconnected so many times toda ?
as I showed we say the TRACERT to be interrupted many times, it was not able to reach the server and the same for a RDC