Tumblelog by Soup.io
Newer posts are loading.
You are at the newest post.
Click here to check if anything new just came in.

April 10 2014

February 13 2014

System update time!

Starting today, this weekend, we will try to migrate soup.io to an updated version of the Ruby on Rails platform (one of the programming frameworks soup is using).

This has nothing to do with new features or anything like that though - it's a purely technical matter that has been a problem for a few years now ("technical debt", for those in the know).

But it's the most important step to bringing you a better experience, new features (beyond the smaller ones which we sneaked in there in the past half year or so) and pixie dust. So bear with us, we've put a lot of time and energy into it the past months, and we're reasonably sure that this can work ;)

We'll try to do it gradually, so if there are any deal breaking bugs, it should affect only a few, randomly selected users, and to those that have been chosen, we salute you! Also pity. But mostly, salute.

We'd appreciate if you could voice any concerns/observations/bug reports with a post @kitchen, or, if even that doesn't work anymore, on our getsatisfaction page.

Thanks a bunch in advance for your patience.

Let's do this thing.
Reposted bysobernaichapertureFate46hexxeMissPunchlineoopsiakrandomuserserenitedunkellichtbesenthor7ostechlingeroutofmyheadqbbaartekinlovegirljulia2999malinowepolejohnfriskci1024banhminhhuonghackyourhealthbrieebankruptcydarwin

August 24 2013

Intermittent crap-outs the next 4-5 days

The database maintenance that we scheduled yesterday morning didn't work out - we had to abort after we've overrun our schedule by two hours. The residual effects of the database recovering from an aborted operation were noticable throughout the better part of the day - which means, we're not trying it again in such a one-shot maneuver.

But since this is neither an unimportant nor a small change to the internal structure of the db, we simply have to do it. We're now opting for the drawn-out method, which means that over the next few days there'll be short periods (10-60 seconds) where soup will work slowly or not at all. But there shouldn't be any bigger problems, so I'd say after a reload or after waiting for a minute, you should be able to continue using soup.

Please understand that we do not see this as normal behaviour and that we want to get rid of such inconveniences for the future (actually fix the bugs), and that's why we're doing this.
Reposted bynaichSeventhxalwonkoBananaRamafrittatensuppesanata-tolakwulfymprhelenwu007carlopesW03callyourleadsprabodh-mehta-gembelTunhuaDaiLoant-bagteppo-vuorijuuriLeMightyMustache

August 22 2013

Scheduled maintenance Fri Aug 23, 4:00am - 06:00am UTC

We'll be doing some database work and can't have you around for it. Thank you for your patience.
Reposted byMarie-EdwigeSpecies5618diptokumarcheapnikef14cps3zqformfillingprojectsfifabusbb15tibianewwillygokilyahoosupport4

June 05 2013

Maintenance window June 6th, 01:00 - 08:00 CEST

I'll be doing a little bit of database maintenance, which means soup will be offline during that time.

I hope it won't take the full reserved window though ;)
Reposted byFreXxXchybakpiszpotrzasknerdanelvolldosteveryonemarsjaninzmarsareduxjanisozaurfadenbagathaableedingmekapitandziwnymakemelaughsheikWoltDiZZneyAdminTongdaiNhaDatcheapnikesnoopyboxaltadentist

March 27 2013

Downtime, groups and assets

Since someone thought it a good idea to mess with critical infrastructure from a remote location without proper precautions or access to the datacenter, soup was offline for a few days around this last weekend. We're sorry about that.

While some of the fallout has compromised a few internal services, the user facing portion has now been fixed, and lo and behold, even improved, by re-activating three servers that were left inactive after the last cable management party at the datacenter.

Thanks to the guys at Nessus, who, as always, responded timely and professionally to our requests for assistance! If you're in need for your own domain name, they have the most comprehensive product in this area called Easyname - be sure to check it out.

In other news, you may now be able to open the group "find and create" page again due to the new found capacity allowing us to extend the maximum connection time. The page will still take ages to load (there's a lot of info being fetched from the database and processed for it, so we'll see how that goes), but it should eventually show up (probably better chances during the day than the evening) and allow you to create new groups. Isn't that something! edit: No, it's not, groups still can't be created. This seems to be a bug in the software that just wasn't discovered since the group page never loaded... m(

Finally, some images went missing when our amazon contract ran out and nobody removed an asset restauration script that then proceeded to overwrite perfectly good images with zero byte files. There shouldn't be too many affected, but if you find any, just re-upload them from their original source. If they are older ones, they should still be around on backups and may turn up within the next weeks. If it's something you truly, truly can't live without and can't get back from other sources, leave a comment to this post with a link, and I'll see what I can do.
Reposted byFreXxXrzrnerdanelbrightbytemarsjaninzmarsastraycatawezonePstrykmanfredxoxoplatoonaperturetowserabluberzahlSirenensangRESERECTTORphintechjotbemicha06dekhabarakhl3viathanschottladenxcwajdaxnaciahasan203susu123520dontmakemeangrymaroko41strip697

August 03 2012

Asset issues due to hardware problems

We are currently experiencing some problems with the main asset cache which got unresponsive. We will try to get it back online as soon as possible, but we have to wait for the team of our data center to have a look at the physical machine. In the meantime, we'll try to get a drop-in replacement ready. We're sorry for the issues.

Update: We now have a fallback asset cache running. Due to the extreme load the asset backend server is currently under (Since the new cache is currently almost completely empty), some images may still fail to load, but that should shortly decline.

Reposted byathalisEusterskasessitasobereveryoneSpecies5618ravhairinmystechlingerexiTamahloelsenkhabarakhjodynaalepetitprinces3itemslitikarchiorab4bluecookiepreposterousheytveryblueberryveryblueberrymisteroram01cupcheapuntook520Easecluesunshine365albertoukm556shengwei1988spee1943wowgoldbookconstructionsoftwaretrimrunescape001runescape001james9durhamMRjavinburdiablo3goldsfun4peak12anjuonecreditswati0104

April 12 2012

April 11 2012

The database wants a small upgrade...

... so we will be offline in ~10-20 minutes (1:15 pm UTC), the whole thing should take around 15-30 minutes max.
Reposted bylajAmericanloverciasteczko7naciasarveshameliachaosbenedictlaqui88toneypanMyraKern11amaarjayquinehopurolimichaelkarlon1wang12345trreligiohluang123langweed59slicesatish

March 27 2012

Scheduled downtimes for hardware maintenance

We will do some hardware maintenance today (28/03/2012) from 00:30 CEST to 02:00 CEST. Most of this should go unnoticed by you, however we probably won't be able to prevent a few downtimes of about 5 minutes within this timeframe. We will try to keep them as short as possible.

For those of you who care, we will improve the cabling in the server rack, move some servers around and reinstall old unused machines. These will most probably then serve as frontend servers for soup.io and/or as backend workers to deal with the feed imports.

Reposted byDerOrwischerfaselselleriethatsridicarusniaachniaachwyszukajwgb79michaelkarlon1wmixsoup

February 09 2012

Scheduled hardware maintenance

We will do some much needed hardware maintenance and disk upgrades today from 10:30 GMT to 12:30 GMT. We apologize for the cold turkey.
Reposted bykmshen97organicskincare01myasiatradeskincareswietlanaarianaswampaneczkaaauuukakiafrancispryorwilmawilmadomkaalonleygirlmalinka4444

December 26 2011

6815 2da6 390
Dec 28

28C3 Soup meetup

15:00 BCC, Raum A04 (beim Hackcenter, neben dem Workshopraum) Add this event to a calendar application
Get shirts, buttons and stickers. Scroll Soup. Complain to @c3o & @lutoma. Meet people (optional).
Reposted fromkitchen kitchen

November 03 2011

Downtime for DB maintenance on Monday

In order to prepare our database for some upgrades in the near future, we'll have to take the Soup database offline for a few hours.

So in the morning of Monday, the 7th November 2011, you won't be able to access Soup from 04:00 to 08:00 CET (that's 03:00 to 07:00 GMT).

We're sorry for the inconvenience.
Reposted bydeinneuerfreundlajsmiNehaleniafadenbmrmagpiepillank8agiennymajkeymisterpetergriberjbeanSol3kchybanieAluappartydomainchybanieTheSieradzkixxqqzzaatymbarkdomkaadar-elsafwa

September 15 2011

Help Wanted: Soup.io/me is hiring

Can you create rich multimedia visualizations with JavaScript, build bleeding-edge HTML5 frameworks or scale popular webapps?
Do you want to work for a freshly-funded, developer-driven startup with 100k+ users that's about to make the jump to Silicon Valley?
Reposted fromc3o c3o

June 24 2011

Next round of data Sokoban!

After we finally have the new file servers under control - can you feel it too? - the next bottleneck is running amuck (gotta catch 'em all!): the database, which is under extreme duress due to people scrolling faster now... who said more bandwidth and swift images would be a good thing again?

So, @terrorobe and me are going to update our database tomorrow, saturday ~10:00 am CEST and also put a new machine in place. If all goes smooth, it should be over somewhen in the afternoon. However, an exact estimate is difficult.

Please watch out for updates on twitter and honor us with some more of your patience. I expect we can reward you with better performance especially during peak hours.
Reposted byDreamingOutLoudmisterpeterprnsxduuulijwanakarafessedodessatllscruzhvactraining9domkaa

June 22 2011

Something's still going wrong.

After the recovery a few days ago, our freshly set up file cluster crashed again (no, no lost files) when we tried to expand its capacity. Now everything is pretty messed up, yet not unrecoverable. I hope I can fix most of it until tomorrow morning, so that soup becomes usable again.

edit 08:28 CEST: That's it for now. If a picture seems to be missing, try reloading it.
Reposted bywaniewskacremembtbeautysandra2011holylandolivewoodrockowaaartztecatymbarkdomkaa

June 14 2011

"Something went wrong"

During the server relocation last night a node of the file cluster gave out, and took it's assets with it.

We are currently running with old and new assets. Right there in the middle is a "donut hole", which some of you might have noticed already because a profile picture is missing every now and then.

We are very sorry about that, but that's life.

What's also life is, that we have offsite backups of everything, so fear not, all is good.

However, the backup replay might take a while: we can do a pretty good job with the new uplink in downloading, but it is a bit more complicated than just replicating a volume. Every asset that's being pulled from the backup needs to be checked if it is still active, downloaded and verified. My guess is, that it will take anywhere between one to five days to bring back everything.

In the meanwhile, if there is a file you really really need now, because it's homework or otherwise extremely(!) important(!!), you can drop us a mail at team@soup.io with the file URI. We will then restore the file for you  manually. Did I mention it needs to be life-or-death important? ;)

Again, sorry for the inconvenience, the next iteration of the file cluster will certainly not be susceptible to this kind of data loss.

edit: those of you who can't open their soup: we are working on that. When your soup opens again, it might not have the design you expect. Please be patient, yours will come back.

edit: the user avatars are now completely restored, which should lead to better performance right away.

edit: the last two days we were very invested in setting up the new file cluster, which is now mostly done, and is still importing files from the old server and the backup. Over time, speed will improve and most important of all, it is built very redundantly. Yay! No more inconvenient data loss in the future.

Also, since we now have more available space at our hosting facility, some other long planned improvements will be implemented over the next days and weeks, all targeted towards acceptable speed and a little bit more stability ;)

Btw, did anybody notice that this little trashcan in edit mode looks like a crown? ... I think I need to go to bed now.
Reposted bymisterpetersoberwandi02mydafsoup-01deepthoughtbigbear3001penpenKarolina22elpollodiablomihoshiwhereistheguruoelsenelvisariasmelenk89rockowaawebtasarimidomkaaautolening

June 12 2011

Bandwidth upgrade, server migration

In light of your bottomless hunger for bandwidth - we now (again) saturate our uplink for more than 12 hours every day - we will move some servers around tomorrow afternoon. This will result in a downtime between 4-6 hours. We estimate the outage to start at 12:00 CET (8:00 a.m. PST) at the earliest and finish at ~20:00 CET (2:00 p.m. PST) at the latest.

This measure will not only bring a new and faster uplink, but also enable us to upgrade some critical infrastructure in a week or so, thus making soup's speed tolerable again ;)

As always, during (and after) the outage, you can find us on twitter for any news and updates.
Reposted bymisterpeterbrightbyte02mydafsoup-01victarionvonxperenoxpenpenkawaipapierosyxxooduuuli

January 24 2011

Coming soon: Server upgrade, new version of Soup

We know that Soup has been slow as molasses for the last few weeks – the load time has quite literally been off the charts. One of the reasons was a lack of available bandwidth. We've now added a second uplink, and the bottleneck has shifted to a lack of processing power – both database CPU and application servers – causing some short downtimes over the last few days.
The good news: We've already ordered the parts and will be implementing a comprehensive hardware upgrade at the beginning of next week.

In even bigger news, as we've hinted at in the past, we're preparing a relaunch of Soup.
In the first version, it will focus on improving the way your Soup account looks and on giving you more control over your content, to sort and filter it and display different types of content in different ways. Changes to the reading experience (friends/everyone) will follow later. Developers will have new opportunities to work with Soup data and change the way Soup works.
Within this week, we will start sending out invites to a separate alpha installation of the new Soup to a first small batch of users. As we react to their feedback, improving the user experience and fixing bugs, we'll gradually invite more people, ultimately upgrading all current Soup users to the new system within the next months.
We can't wait to see what you think and how you're going to put the new featureset to use!
Reposted bydeinneuerfreundelpollodiablotomenunfoldedfpletzatomanfaselspiereksoberhdsCattalizersstefaniabrightbyteekeliasOsorkoniusconsumrsofiasalviondlsanojmoonsterigi666daniovolldostmr-absentia02mydafsoup-01awezoneleonnero-NeoZaubertrankgriberpsychobkrolikmawDiviuschybanienylkimolotovjaen-ni-rinsellerieabldziedzicpruskiLukasYorkvoid37Xavaxcremefrequenzwasserle-dernier-bon-samaritainacidkrannixpoolunkayarbuziatkosvslunceMeinSueppchenkampfbrotkadreichickinsoupriotiberivylPapsTHoazlwolfssuppejthanAluslawschokobonszupa-witadersksowaleonellezshothouvermillionlovebartoszpiatkowskisomethingpenpentorikocristiancarickycharlesdomkaa

September 29 2010

5104 5e6c 390
Control filtering of Everyone and Friends streams

First, thanks for all your feedback on the NSFW filtering issue. There are a lot of opinions and people are obviously split on which way we should go. So we've decided to introduce a way to disable filtering.

If you'll look at the top of your friends/friends of friends/followers and everyone stream, you'll notice a new checkbox called "Show potentially NSFW content". If checked, Soup.io won't filter content marked as NSFW from your streams.

Please note that this setting is currently stored in your brower's cookie, and defaults to "off". We will store it in your user settings if we're happy with how this feature works.

We know that this isn't the best solution, and that e.g. per-post filtering would make more sense. We will keep working on it and eventually implement a better solution, but until then you at least have the option of choosing for yourself wether you'd like to filter or not.
Reposted bydeinneuerfreundgriberfinkreghraniSigalon02sicksinflower314domkaa
Older posts are this way If this message doesn't go away, click anywhere on the page to continue loading posts.
Could not load more posts
Maybe Soup is currently being updated? I'll try again automatically in a few seconds...
Just a second, loading more posts...
You've reached the end.

Don't be the product, buy the product!

Schweinderl