BitTraffic Status

Zeus’ long running service came to an abrupt ending this morning when a data-centre tech prematurely disconnected it just after I’d made a backup but, hey, all is good. We decided to replace Zeus (UK2.net ~ London, England) with Pandora (ovh.co.uk ~ Paris, France) because it made better financial sense and, would you believe it, Pandora is also much faster.

The end user will not notice any massive difference in the service provided by BitTraffic although the new server will be faster to respond to requests, reverse DNS requests will be processed faster and customers can now make use of IPV6 addresses if they wish. Pandora is running FreeBSD 7.0-Release, the latest and most stable FreeBSD release available.

As mentioned thanks to OVH BitTraffic now offers IPV6 functionality including the availability for clients to run websites, ircd servers and such like on the latest generation of Internet Protocol. For more information please contact us.

BitTraffic’s name servers have been updated so as long as your domain’s nameservers read ns3/ns4.bittraffic.com your website will work just fine.

Many thanks for sticking with us, we do appreciate your business.

Chris
Bittraffic admin

At present we are experiencing problems with our openvz system thats houses pegasus & all vps systems. We are in constant communications with datacentre to resolve this issue asap.

The situation now stands that system might be down for a few days as datacentre has advised that we migrate to an alternate box. They can make a backup of our system which can be restored to new box.

Unfortunately this downtime has happened whilst Stalker is on his holidays so we cannot relocate any clients to alternate boxes yet but will do our best to contact Stalker to come to some arrangement for clients. If Stalker cannot connect to internet whilst he is away for any reason he will sort something for downtime asap upon return which is 9th June. I have messaged him & he is aware we are having problems & I’m sure he will do his best to connect to help asap.

I am sorry for any inconvenience caused by this downtime & will do my best to get it sorted asap for all clients.

We have taken steps to ensure that servers that are of highest risk to large apache loads and/or floods have been given a helping hand in dealing with such issues. Apache has been limited to the amount of resources it can hog and whilst allowing this service a relatively small amount of system memory and cpu-time we have ensured that genuine http requests will not be disallowed and that access times have not been slowed down.

Our aim for this small project is to increase all-round system firmness, not to stop genuine http requests from being processed in the usual short amount of time. However, if any user notices downtime, particularly of http, we urge them to contact us with the server name, time, date and length of downtime so that we can maintain correct thresholds for resource limiting.

Chris
Bittraffic Admin

The NIC card inside Athena began slowing down and was replaced, thus forcing a system reboot.

Service is back to normal.

Chris
Bittraffic Admin

All servers are now running the latest versions of Apache, PHP and MySQL.

We have also added a module to apache to help keep a particular system stable in the event of a DDoS [distributed denial of service] attack.

Chris
Bittraffic Admin

In a continued effort to bring our customers the latest available software to ensure as smooth a hosting journey as possible for everyone we have just upgraded the majority of our servers to run the latest versions of Apache httpd (v 2.2) and Php (v 5.2). I trust that these versions are backwards compatible with any older websites/scripts that our customers are hosting and thus will not affect performance negatively however if any user does experience any problems please post a support ticket at http://support.bittraffic.com and a member of the support team will respond ASAP.

Thanks,

Chris
Bittraffic Admin

We’ve fixed the errors with compiling. The problem was related to hardware as we suspected because with some new memory the box is now running normally again. I’d like to thank all of our customers for being patient and hope that little convenience was caused.

Chris
Bttraffic Admin

Dear BitTraffic Customers,

Users on the Hercules server are having problems compiling applications, we’re experiencing “Signal 11 - segmentation fault” errors. This fault is most likely related to hardware (e.g. overheating or worn out memory) but we havn’t yet identified the exact problem. The problem began occurring around 24 hours ago and technicians are working on the issue as we speak. Unfortunately it is likely that a reboot will be needed to restore full working order on the server as the replacement of hardware is looking to be quite likely.

Here’s a typical error we’re getting from the compiler;

hercules# make
cc -c -O2 -fno-strict-aliasing -pipe -Wall -DHAVE_CONFIG_H -DCONFIG_FILE="\"/usr/local/etc/healthd.conf\"" -I. -DINET6 getMBinfo.c
getMBinfo.c: In function `InitMBInfo':
getMBinfo.c:71: internal compiler error: Segmentation fault: 11
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://gcc.gnu.org/bugs.html> for instructions.
*** Error code 1

In all of my years working BitTraffic i’ve never come across this particular error so if anyone has had any experience with it give me a shout at chris “AT” Bittraffic.com and i’d be grateful.

I’ll be sure to keep you updated on this issue and apologise for any inconvenience caused during this forced maintenance period.

Chris
Bittraffic Admin

Dear BitTraffic Customers,

We apologize for the issues related to FreeBSD 6.1. We have now upgraded all systems to FreeBSD 6.2-RELEASE-p2 the latest stable kernel available. Its been running excellent. We apologize for our decision to use 6.1 without having further testing and preperation. The issues that we had were uncommon and were related to IRCd’s only. They were not well known and the issue was not easy for our techs to troubleshoot, and was not even well known by OS guru’s around the world. The new kernel has patched the issues associated with the IRCd problems. We appreciate that you all have been able to bare with us through these kernel issues.

Thank you for understanding,

Dear BitTraffic Clients,

At approximately midnight MST hercules.bittraffic.com running the kernel FreeBSD 6.2-RC1-p1 crashed proving unstable and unreliable. At that time we attempted to upgrade the kernel using standard procedure that was successfully used on both stamina.bittraffic.com, and zeus.bittraffic.com. We upgraded the kernel to FreeBSD 6.2-RELEASE-p2 and failed. After the final configuration changes were made and the system was restarted it failed to boot. We then submit a ticket to the datacenter regarding the issue. Further investigation revealed the kernel problem. The kernel was replaced and updated with a working FreeBSD 6.2-RELEASE-p2. This update was critical the same as the first attempt with p1 and refers to the same issues. The system hercules.bittraffic.com should now be stablilized with the newest available stable release. We greatly apologize for the disturbances and are confident that the issues are resolved. If you have any questions or concerns please do not hesitate to contact us.

Thank you,