Currently use the Groupdocs API for managing files on our site, between 8am and 8.30am we had issues loading any files via the direct viewer (viewer simply returned error getting content) and downloads were only returning empty files (0KB), looked like an issue on your end (we have had this a few times in the past) and when I contacted support around 8.30am it was resolved but I have been asked to get a report on why this issue happened.
Can someone provide me with reasons why groupdocs was unable to service files between this time on 04/08/2015?
Hi again, I have now conducted an investigation on our sites and this actually seems much more severe then originally noted, since 9PM 04/08/2015 a significant amount of files attempted to be uploaded via the groupdocs api are now stored as 0KB files… it seems like the issue was not just a display issue of current files but actually effected the uploading of files as well.
This is a huge problem as the window of ‘downtime’ is now 9pm - 8.30am and we are left with lots of corrupted uploads by our customers/clients, I will definitely need a serious response to this as no code was changed on our end and the API is used on three sites which use different code and they all have been effected.
Would appreciate a relatively prompt reply on this.
Our product team has investigated your issue and found out that in this time interval was scheduled new update of our Cloud API for our service and in this time interval our service was unavailable. We apologize for the inconvenience.
Could you please try again and notify us with the results.
Thanks for the reply however this isnt really a sufficient response. The API is working correctly now (as of 8.30am), the issue is the API was not returning any errors and simply confirmed files were uploaded but they were not uploaded correctly as most of them are now simply 0KB corrupted files so somewhere in your API it failed to do the actual transfer.
I am at a loss at how it was acceptable to have this scheduled in and as far as I am aware we have not been notified of any planned updates. Our sites requires reliable storage and the fact we now have lots of corrupted files from many of our users is a rather dire situation especially as this effected uploads over the span of 11.5 hours.
Is there anyway you can recover these files or are they completely unrecoverable?
Sorry for misunderstanding. Looks like the issue was not with deploy. Product team is investigating the logs to try to find out what was wrong with your requests. About the files, we afraid that they cannot be recovered and they should be re uploaded. But we will notify you if we will have any new information.
Could you please clarify, did you get the issues with this account - lee@ico3.com, or you have other accounts? You see, our product team have checked this (lee@ico3.com) account, but looks like it's pretty new and trial account. Also please specify which API server you are using https://api.groupdocs.com or http://eu-api.groupdocs.com ?
Please come back to us with these details and we will be glad to help you.
Well I was told this was being investigated… the issue was not ‘resolved’ as we have had to manually get the files reuploaded and verify every file for 11 hours worth of corrupted uploads due to no one notifying us of updates to the back end which is apparently the cause?
I am waiting for an incident report, why was the backend broken but reporting back response codes as correct, 11 hours is far too long as we were not even able to identify an issue because your api was accepting uploads but simply failed to pass them across (which is all on your end, not our code).
You had scheduled maintenance a few days before this issue so why wasnt it upgraded at that point, this is the main issue.
Unfortunately there is yet another issue as your whole platform seems to be down since 13.30ish, someone else in the office is most likely trying to contact you but if you could look into why this is down that would be great.
We are very sorry that you have the issues with our service.
Could you please share with us, which Api methods you have used and we will investigate them with our product team. We will try to implement some protection from such incorrect responses from the server for these methods ASAP.
We are using the GroupDocs Viewer with Google storage, on the PHP API class. The account is with our client TDS (Alan@tds.gb.com)
The issue as described above has reoccurred again today, between 11:05 GMT and 11:45 GMT all files uploaded across all 3 of our systems were returning that the file had been accepted & uploaded correctly, but when trying to access these files found they all showed as 0kb.
When speaking with Pavel he said this could be caused by a “space issue” that “wasn’t spotted fast enough”.
As the API did not return any failure messages we were unable to notify our end users and as this was within our peak hours, have now had to pause normal business procedures while we contact hundreds of public users to re-upload their files.
This is in addition to the hours downtime on Monday 24th August between 13:30 and 14:30 GMT which caused us major disruption and for which no root cause was ever determined.
Can you please urgently let me know what steps you have taken to prevent similar outages in the future as we cannot afford for any recurrence in the future.
We discussed the issue with Platform team and they notified us that there was very high activity on your account so temp folder was overflowed. Temp files of your API activity for this month were about 150Gb. Platform team configured special tool to remove temp files daily, so now there should be no such issues with free space. Also we asked the team to investigate how we can improve errors handling for files uploading. They promised to investigate what they can do for this. We hope that they will be able to add error notification for the case you mentioned.