Answered by:
Recurring corrupted database

Question
-
On a regular basis an Access 2013 gets corrupted and Compact and Repair fixes the problem.
Is there a way to investigate the reasons that are causing the corruption?
- Edited by JohnPapa05 Saturday, September 1, 2018 12:28 PM
Saturday, September 1, 2018 12:27 PM
Answers
-
Hi,
the culprit most probably is
https://support.microsoft.com/en-us/help/4100403
OS Build 17134.81 of 2018-05-23
There was some hope that this later update would have cured the problem, but several users have reported that it didn't:
https://support.microsoft.com/en-us/help/4340917
OS Build 17134.191 of 2018-07-24If you are not on a deferred update channel you should have gotten both.
Normally it would not be advisable to disable "Leasing" i.e. Windows caching as this may cause bad performance. However so far this workaround is the only reliable solution for the problem.
Karl
http://www.AccessDevCon.com
http://www.donkarl.com- Edited by Karl DonaubauerMVP Tuesday, September 11, 2018 8:22 PM
- Marked as answer by JohnPapa05 Thursday, September 13, 2018 1:23 PM
Tuesday, September 11, 2018 8:20 PM
All replies
-
What's your setup?
What is the size of the database file?
Is the database Split?
Does each user have their own personal copy of the Front-End on their local PC?
Do you establish a persistent connection at the startup of your db?
Are you running over a wired LAN connection, or are you using wireless or a WAN setup?
Does the corruption occur randomly, or is this reproducible (a specific action causes this)?
Daniel Pineault, 2010-2017 Microsoft MVP
Professional Support: http://www.cardaconsultants.com
MS Access Tips and Code Samples: http://www.devhut.net- Edited by Daniel Pineault (MVP)MVP Saturday, September 1, 2018 12:51 PM
Saturday, September 1, 2018 12:50 PM -
Here are a few articles of best practices and setup principles, be sure you are following them.
- http://www.devhut.net/2017/04/09/setting-up-an-ms-access-database/
- http://www.devhut.net/2017/04/20/access-best-practices-and-troubleshooting-steps/
Also, what's your network like? How fast and stable is it? Access is heavily reliant on it's network connection, so if yours is less than optimal then that could be the entire issue right there.
Daniel Pineault, 2010-2017 Microsoft MVP
Professional Support: http://www.cardaconsultants.com
MS Access Tips and Code Samples: http://www.devhut.net- Edited by Daniel Pineault (MVP)MVP Saturday, September 1, 2018 12:55 PM
Saturday, September 1, 2018 12:55 PM -
Daniel thanks for your questions. Here are the answers:
What's your setup?
Db on separate PC and 7 front ends
What is the size of the database file?
About 80Mb (relatively small)
Is the database Split?
Yes
Does each user have their own personal copy of the Front-End on their local PC?
Yes
Do you establish a persistent connection at the startup of your db?
Yes
Are you running over a wired LAN connection, or are you using wireless or a WAN setup?
Wired
Does the corruption occur randomly, or is this reproducible (a specific action causes this)?
It is random
Saturday, September 1, 2018 1:12 PM -
Had the network checked by two network people and they say it is fine with fast access.
I have a second Access db (same software) running on same network with 2 FEs and there is no problem.
- Edited by JohnPapa05 Saturday, September 1, 2018 1:16 PM
Saturday, September 1, 2018 1:13 PM -
"I have a second Access db (same software) running on same network with 2 FEs and there is no problem."
Well, that give us a path. If it doesn't happen in other DBS, then it is most probably a DB issue.
Does it happen when only one person is using the DB?
What about creating a new blank DB and importing everything and seeing if that helps.
Daniel Pineault, 2010-2017 Microsoft MVP
Professional Support: http://www.cardaconsultants.com
MS Access Tips and Code Samples: http://www.devhut.netSunday, September 2, 2018 4:40 PM -
The problem is random. It is not user specific.
It crossed my mind to copy the backend db into a new db. The FE is used by 100 other people elsewhere without a problem.
It may be pc specific. One faulty ethernet connector would do the trick.
Monday, September 3, 2018 5:03 AM -
Hi,
you didn't tell the error messages etc. but if it's the backend that gets corrupted and if Windows 10 is involved on the server or any client then you might have the most frequent corruption problem these days.
It is caused by the Windows caching and depends on the build number(s) of Windows 10. Which builds do you have or at which Windows update channel are you?
The workaround for the problem is to disable caching in the registry. Daniel has more on the drama at his website.
- Edited by Karl DonaubauerMVP Monday, September 3, 2018 5:57 PM
Monday, September 3, 2018 5:57 PM -
Hi Karl and sorry for the delay. I missed your reply.
Which Windows Build are you referring to? In a specific network we have 10 pcs.
In this network we changed the Switch and the software runs faster and the problem has not appeared.
Is it advisable to disable caching anyway?
Monday, September 10, 2018 1:38 PM -
Hi,
the culprit most probably is
https://support.microsoft.com/en-us/help/4100403
OS Build 17134.81 of 2018-05-23
There was some hope that this later update would have cured the problem, but several users have reported that it didn't:
https://support.microsoft.com/en-us/help/4340917
OS Build 17134.191 of 2018-07-24If you are not on a deferred update channel you should have gotten both.
Normally it would not be advisable to disable "Leasing" i.e. Windows caching as this may cause bad performance. However so far this workaround is the only reliable solution for the problem.
Karl
http://www.AccessDevCon.com
http://www.donkarl.com- Edited by Karl DonaubauerMVP Tuesday, September 11, 2018 8:22 PM
- Marked as answer by JohnPapa05 Thursday, September 13, 2018 1:23 PM
Tuesday, September 11, 2018 8:20 PM -
Karl,
I will see if it happens again and will report back to the forum.
Until then, I will assume this is the answer, but as it was mentioned the network must be stable and fast. In one case we changed the Network Switch and it seems to have fixed the problem. Fingers crossed.
Thursday, September 13, 2018 1:23 PM