Search results

Just some statistics, if you want :) SQL-Database is about 50GB of Size (as pure SQL txt statements) we are currently doing 600Querires / Second on our database currently we have 3 cluster-nodes with the other each doing like ~100 Queries/Second (and doing backups etc. op top) Currently...

*sight* php crashed & i didn't get any notifications -.- going to prepare a better solution

As you may have noticed, we were offline. I'm sorry for the unplanned offline time; we had an issue with the database; and i fucked up the debug-config; so i searched for the error longer than may have been necessary. but a (very) few hours of sleep and it worked out. also just for a...

if something went wrong; everything will ... letsencrypt didn't renew ~fixed. ~

not so really sure, what happened ... main-system did look ok; no errors in any logs ... reboot of database somehow helped.

Asia hosts are currently down :/ not sure, what happened. DNS should reroute to EU nodes.

and ... the whole site crashed :( they have changed the pattern & ultimately the database crashed. hopefully it's now fixed?

*sight* european (or US) script kiddies are still trying to do shit with us ... Nearly everything is filtered on the frontend-servers; But perhaps there might be some issues; Providers are currently holding up. (which is quite suprising, since they had massive issues in the past) Asia...

oh... going to check. (that explains, why i'm not getting a mail from this watched threat -.-) edit found the issue; there was a NAT-missconfiguration (aka: i forgot to enable the firewall rule)... should now be fixed. edit2: I cleared the mail queue; since we do have a limit, how many mails...

totally rewrote the entire DNS-System. we are now also checking for existing VPN-connections & we are return a random entry in your continent dataset. If you are located on the continent "europe" you will get the EU servers; if you are in Asia, you will get the Asia servers; if you are...

i've enabled the old frontend-servers; we now have 2 with ... questionable uptime. I've set up a new DNS-system, which checks every 30s for offline-servers and changes the DNS answers. I had high hopes, that this will fix the unstable connections in EU. But as it looks like, this is...

update: after the cleanup, the remaining uploader, did exactly the same! WTF ey! -> upload to the hentai manga will be restricted to people with an upload tag! If they are not able to use the correct forum, i don't care anymore. Seems like the part is quite empty, if you remove the 2...

i would suspect directly here: https://www.akiba-online.com/forums/hentai-downloads.35/

You guys did it again ... Because of some random shitty user posting something that is against nr. 9 of "THE RULES" we have lost our best working frontend. It will not come back online; I won't argue with them about something that is clearly shit and obviously CASM. I'm using their services for...

yay; it happened again. we lost one of our EU servers ... DNS is updating.

we are experiencing some strange behaviour the last day. seems like someone is hammering requests to us... resulting in some failsave actions, that sometimes let's the site hang :/ (see the spikes till nearly zero) That should not happen, but i'm not sure how to mitigate such requests hammering...

*Grrr* DDOS attack to one of our european frontend Servers :/ after a long time, i removed it from the DNS - some some flaky connection issues from EU. Gonna check, if i can get some kind of prof. CDN - and not some lowcost do-it-yourself solution.

ah yes. sorry about that :D i forgot to update this threat after the search-index-rebuild was done

Going to check this in a few hours. I need some sleep before messing with elasticsearch

sorry for the offline-time backup broke and the database and the main-system vm wasn't "defrosted" ... not sure why this happend, exactly only on the akiba systems; every other system seem to work without an issue :( als i did get notified, but i was sleeping :(

I've enabled a 2nd - new - EU server. (thanks to @chompy for paying :D) If you experience some issues, please let me know. edit: well ... fuck. added it - got offline :( edit2: oh well ... someone rebooted the system

it's an issue with the EU-Node. I've disabled the EU node - and will cancel it. I've deployed a new System from another provider. let's see, how this one works. (also I re-enabled one of the old-old eu nodes; in the hope, that it will work without hickups) I'm sorry for the ongoing issues...

... i'm looking into a heartbeat for DNS ... and new frontend-systems :( the last EU-Server is hosted on a network that receives DDOS-attacks quite frequently and the connections are somehow halfway dropped, when the protections seems to start :( Thanks to @CoolKevin for pinging me. Monitoring...

sorry for the short offline time. an update wrecked the system (aka: cached CSS didn't match the newly build one) also i did figure out, that one of the 2 asia-pops are not working :( so we are currently down to 1 in asia 1 in eu (vs 2 asia und 3 EU and 0 in US :( ) edit: 2nd asia node is ...

currently there seems to be less issues - but it's not optimal. DNS should now have propagated to everyone. e.g. within EU it's okay-ish within Asia it's okay-ish within the US it depends ... on the east-cost it's okay, on west-coast it's absolutely shit to browse akiba Nevertheless a complete...

the EU frontends are - unfortunately - currently very unstable; i've disabled the shitty ones; let's see what happens. The system does have some issues, sometimes; but the last week was really unstable. Asia frontend has a more than normal latency - not exactly sure why (more than 150ms) :(...

You may experience some slow speeds currently - we have lost our ipv6 endpoint. I've updated the DNS-Servers and removed the Server - propagation may take some minutes. edit: ipv4 should be back to default operation speeds. ipv6 is still offline1

hi, yeah; i didn't found the time yet. So ... do you remember when AWS was down? Well we fucked up the same way - it's always DNS. We had to move some domains around & i messed up the DNS-Server addresses. Since my monitoring only checks IP-Addresses, I didn't get any information in the...

moved one of the EU servers back online; 2nd one is still funky -.-

EU nodes are fucked up again ... so we are switching to Asia only - at the moment.

*sight* this would have been way to easy, right? so ... again sorry for the offline-time. Post-mortem (as of now): Mysql does have an issue with limits in their query and very large datasets. There was someone / more than one? That queries all threats of the forum. With the bug of mysql a...

Database got stuck. The whole system did have 100% load and very high io... Still investigating what happened. sorry for the offline time:

IPv6 is online again. Please report issues, if there are any. Europe frontend is still behaving strange sometimes, but there are not that much alternatives. - and they are stable again for the last 2 weeks. (only 2 small hickups)

ipv6 offline; the upstream-provider did have some severe problems with their BGP... I'm talking with chompy about possible alternatives.

so... i fucked up a config & needed to wait till i got to my notebook again. sorry for the offline-time. Backup-Issue is not yet solved :(

Still not solved <.< Disabled the backup until I’m home from Holliday Also got an error on the filesystem… again … I hate when that happens

Backup fucked up. Only 300kb/s is way too slow. So the typical 1min hiccup was 1h - I stopped it. I must look into this - new - error <.< Sorry for the inconvenience

Not again! We need to find a new provider if this keeps going on :(

Server 2 seems to be stable for the last 24 hours. I will monitor it and return the 1st eu-node back to production… Not sure how to handle this a better way. Some times the network of the provider does very strange things. :/

Im going to puke on these servers … The 2nd one worked flawelessly. I removed the one, who has errors - now the 2nd starts throwing errors … Going to look into this

Hi. Im going to check this out. At the moment we do have an issue with one of the European frontend systems. It’s the same that got the same issues the last time. Perhaps we have to ditch it.

Sorry for the unplanned maintenance; I urgently had to patch the firewall (and in this case also change the SSD...) Systems are back online, going to check. The last script i hacked it working - but nevertheless it produced about 1k Error-Messages per day. If there is a new plugin, i'm going...

Sorry for the upload issues. the backup went wrong and pushed all VMs into the grave with IO-errors :/ seems like a proxmox issue with failing backups. Let's see if we need to change this in the future.

yes. We still have issues with the EU frontends :( they drop the Connection every few minutes to hours... I've removed the most unstable-one from the DNS Status: Moved to new Hardware; Upgrade of Software in Progress 1/?: moved all Data to the new Server [done] 2/?: Database-Backup [done]...

more or less back to the old design (the dark theme is back, too. You need to change to it - on the page scroll down to the bottom and click on "UI.X Akiba") There are still lots of problems. Missing Plugins: force-words; Thread-Thumbnails some plugin is also causing errors when the system...

as you notice ... we are back on the default xenforo-theme. i'm sorry, that this didn't work out like i wanted. Status: Moved to new Hardware; Upgrade of Software in Progress 1/?: moved all Data to the new Server [done] 2/?: Database-Backup [done] 3/?: Database-VM-Backup [done] 4/?: Upgrade...

I think there was a problem with the backup of the database-server. I didn't really thought about chaning my backup-strategy; so every sunday it will do a full backup including a freeze of the system. Timing should match. I'm changing this (as for your info; cluster node 3 is normally used for...

the new database-cluster is working now. Commencing data-transfer.

Well... I need to buy another server; the 3rd database-node is f***ing slow :( until then (probably a few hours) the system is back online on 2/3 of a cluster. As already mentioned above: expect a little bit slower system than what you are expected due to latency. edit somehow we hit a MySQL...

Oh; I can recycle this thread - nice! After a few years on the current infrastructure - we need to move again. The increased pricing and resources we need (more than 1TB of storage, 100GB Database, ~20TB traffic) are forcing us to move providers. I will try to have as little downtime as...