Comments

• mrerenk Member

  12:10PM

  GLWS
• stufently Member

  12:27PM

  Hi
  Can you add many buckets per one order and api please.
• hyperblast Member

  12:43PM

  @systemfreaks said: Canada & Europe locations

  nope!
  no europa atm!
• Obelous Member

  12:46PM

  @hyperblast said:

  @systemfreaks said: Canada & Europe locations

  nope!
  no europa atm!

  The FAQ on their site also says:

  Systemfreaks Object Storage is available in Falkenstein (DE), France (FR), Canada (CA), and USA (US)

  Yet none except Canada show up, lol
• rad964 Member

  1:15PM

  I have offer from systemfreaks from this

  I think it mostly works, I can upload files, their service serve files

  But there is caveats I faced, and for now, I don't know if I'm reason for this or not
  1. I can't upload much concurrently, I constantly faced with 429 slowdown error when I try to surpass ~12 concurrent uploads
  2. I don't see bucket rules work, I have 5 rules for prefixes for around 2 weeks now. They set to remove files from certain prefix older than 1 day. They don't work. File manager not working/helping either. Clients can't delete files either, because you need to list them in prefix, and that times out. Which leaves me out with around 600gb of files I can't delete.

  When first problem is just annoying and fixed with implementation of spillover queue on my side. But second one probably will need to be dealt with at some day.
  I asked support about non-working rules other day, and when their replies was fast, they failed to resolve problem and their last reply was quite strange

  The LLM job is scanning the whole bucket and deletes files 5937546 Objects are still pending deletion, the speed rate is normal for SATA disks, you can either wait for it to finsh or you can create different jobs pet /path and start using the bucket with different path names so the files aren't deleted
  

  At least written by human, judging by miss-spell
• Obelous Member

  1:19PM

  LLM job? wtf
• JabJab Veteran

  1:24PM

  LLM job to delete files?
  based on some easy to implement rules like older than 1 day?

  W T F?
• systemfreaks Member, Patron Provider

  2:13PM edited 2:17PM

  @rad964 said:
  I have offer from systemfreaks from this

  I think it mostly works, I can upload files, their service serve files

  But there is caveats I faced, and for now, I don't know if I'm reason for this or not
  1. I can't upload much concurrently, I constantly faced with 429 slowdown error when I try to surpass ~12 concurrent uploads
  2. I don't see bucket rules work, I have 5 rules for prefixes for around 2 weeks now. They set to remove files from certain prefix older than 1 day. They don't work. File manager not working/helping either. Clients can't delete files either, because you need to list them in prefix, and that times out. Which leaves me out with around 600gb of files I can't delete.

  When first problem is just annoying and fixed with implementation of spillover queue on my side. But second one probably will need to be dealt with at some day.
  I asked support about non-working rules other day, and when their replies was fast, they failed to resolve problem and their last reply was quite strange

  The LLM job is scanning the whole bucket and deletes files 5937546 Objects are still pending deletion, the speed rate is normal for SATA disks, you can either wait for it to finsh or you can create different jobs pet /path and start using the bucket with different path names so the files aren't deleted
  

  At least written by human, judging by miss-spell

  I believe I managed to locate your ticket based on the description. It appears that your bucket contains a few million objects, and you previously contacted support to help delete the files. We set up a lifecycle rule for the specific paths you wanted to clean up.

  The current issues you are experiencing with the bucket appear to be caused by your application. In short, you need to:

  1. Remove deleted objects from your database to avoid excessive 404 responses from the API.
  2. Fix your health checks, which are still making requests for objects that have already been deleted.

  Once these two issues are resolved, we can review the 429 errors and investigate them further.

  Since this is a public reply, I cannot share more information here. However, if you reply to the ticket regarding issues 1 and 2, I will be more than happy to help you find a solution.
• systemfreaks Member, Patron Provider

  2:18PM

  @hyperblast said:

  @systemfreaks said: Canada & Europe locations

  nope!
  no europa atm!> @Obelous said:

  @hyperblast said:

  @systemfreaks said: Canada & Europe locations

  nope!
  no europa atm!

  The FAQ on their site also says:

  Systemfreaks Object Storage is available in Falkenstein (DE), France (FR), Canada (CA), and USA (US)

  Yet none except Canada show up, lol

  Out-of-stock locations are currently hidden from the order form. Please try again later or contact us to check availability for your preferred location and storage capacity.
• Obelous Member

  2:19PM edited 2:50PM

  did you accidentally paste your LLM prompt? lol

  
• systemfreaks Member, Patron Provider

  2:23PM edited 2:29PM

  @Obelous said:
  did you accidentally paste your LLM prompt? lol

  

  Copy pasted my "Note" since that was missing from the first post, and i had it written on the leave a comment box , but good catch
• rad964 Member

  2:38PM edited 2:40PM

  @systemfreaks said:
  1. Remove deleted objects from your database to avoid excessive 404 responses from the API.
  2. Fix your health checks, which are still making requests for objects that have already been deleted.

  I will check of cause
  I also don't see how your answer suppose to help situation about not working bucket rules. The fact that I have millions of object - doesn't whole point of S3 was to handle millions of objects?
  P.S. I don't get how 404 requests are affecting something
  P.P.S I will write to support later, hope LLM job do better this time
• systemfreaks Member, Patron Provider

  2:46PM

  @rad964 said:

  @systemfreaks said:
  1. Remove deleted objects from your database to avoid excessive 404 responses from the API.
  2. Fix your health checks, which are still making requests for objects that have already been deleted.

  I will check of cause
  I also don't see how your answer suppose to help situation about not working bucket rules. The fact that I have millions of object - doesn't whole point of S3 was to handle millions of objects?
  P.S. I don't get how 404 requests are affecting something
  P.P.S I will write to support later, hope LLM job do better this time

  I can't share more information publicly about your bucket, which is why my answer may seem incomplete. There are additional details specific to your bucket that I can't disclose in a public discussion.

  I'm happy to assist you further within a support ticket, where we can discuss the bucket and its configuration in more detail.
• johndeo983 Member

  2:52PM

  Still using minio for the backend?

  Paid plan or unmaintained?
• systemfreaks Member, Patron Provider

  2:58PM

  @johndeo983 said:
  Still using minio for the backend?

  Paid plan or unmaintained?

  Minio in production
  rustfs is also being tested on replication cluster