New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Comments
@vld
few weeks ago francisco talked about getting rid of buyvmstatus because there are lots of false positives and newer nodes aren't added to it. @vld is the maintainer and the owner of the site. He also mentioned that he haven't changed anything in about an year. So yeah..
I was talking to Fran the other day. He said there were issues related to the IP addresses that were referenced on that list being wrong. The IPs changed for some reason that I do not know. A new server status list might be in the works? @Francisco
We asked VLD to stop running it while he was basing the results on ping. With us offering filtering it isn't uncommon for people to smack at hops before the IP in question (namely the node IP as well as router IP's).
In the end when people did that autonull would kick in and nullroute the attacked IP. This didn't affect a users connectivity but that didn't matter. We would get 30+ cases a month where users would demand SLA based off buyvmstatus, even though their own VM would show 100+ days uptime and their own personal monitoring showed no problems.
The last straw was when Aldryic had a 6 reply argument with a client angry because, according to buyvmstatus, Stallion was down all month. In reality, buyvmstatus had an old IP for manage recorded and we didn't bother binding that IP to the stallion 2 server.
At some point i'll put together an official monitor that will display more information than buyvmstatus did (loads, uptime, etc), as well as base information from inside the networks, not just externally.
Francisco
So no more four OpenVZ nodes downtime at once? :P
No more false positives in general
We use nodeping for our own internal monitoring but it's sometimes hit and miss as well. We still use it as a basis for SLA's, though.
Sometimes it's just cogent being dumb. We've had nodeping report the core router down yet there is no nulls or anything. Thankfully we're moving cogent to peering-only on the west coast in the coming months.
The incorrect IP's weren't a fault of @vld or anything. He did give Anthony the ability to update as well as add/remove entries, but we were always forgetful. It was still missing a bunch of entries in New Jersey as well as a few KVM's in LV.
Francisco