| Subject | Firebird stop responding |
|---|---|
| Author | mivi71dk |
| Post date | 2009-12-15T08:05:04Z |
Hi
We have a server running firebird which periodically stops responding.
This server is running Debian and Firebird 2.1.3 SS (just upgraded from 2.1.1 yesterday). Its has 12 Gb RAM.
On average it has 100-150+ connctions.
At least once a week the firebird server stops responding. Users already connected freezes and no new users can connects.
The easy way to solve this problem is to shutdown the database and restart it. Then it all runs again.
This happen twice yesterday.
When this happens the gab between transactions can be 5000 or it can be 250000.
The clients connects to 7 different terminal servers. From here they connect to the firebrid server.
On thesse 7 terminal servers a lot of other clients connect as well, and the connect to other databases. As it is now we have:
- 1 server running windows and Firebird 1.5 (running 14 Firebird DBs)
- 2 servers running windows and Firebird 2.1.1 (some 10 Firebird DBs)
- 1 server running windows 64 and firebird 2.1.1 64bits (some 9 Firebird DBs)
- 1 linux server (the above) running Firebird 2.1.3 and only 1 DB
All the windows servers with the different firebird versions runs just fine. They never stops responding.
Outside our house we have a lot of setups running. Some uses they own Terminal Server, A firebird server and connect to it as we do.
Others have servers locally and have clients connect locally.
Others just have one machine with all on this.
They use Firebird 1.5 and Firebird 2.1.
None of them has this problem.
The problem is only with the linux server. And this is our biggest DB. Its almost 30 Gb in size now.
In the firebird log we have a lot of these entrys
DBSRV06 (Server) Sun Dec 13 10:02:56 2009
INET/inet_error: select in packet_receive errno = 9
About a year ago we moved this DB from a Windows server because of this exact problem. But we never solved it. It might not happen as often on this linux server, but it happens.
We have come to a point where we do not know where to go from here.
Our client is not happy anymore. They have asked, what we can do.
My colleques has asked if switching to another DB would help (they have mentioned mySQL, Progres, MS SQL and so on). But all of those requieres a lot of reprogramming.
My colleque has been googling some on this problem, and has seen that others experience the same problem.
As of now we have come to the conclusion, that somewhere in Firebird there must be a problem.
But we have no idea what to do now.
If we do not find a solution, is there someone out there that we can pay to help us fix this problem?
I would hate to switch away from Firebird. Besides this customer DB we do not experience any problems, and Firebird runs just fine.
But on the other hand we cant afford to loose this customer.
Regards Michael
We have a server running firebird which periodically stops responding.
This server is running Debian and Firebird 2.1.3 SS (just upgraded from 2.1.1 yesterday). Its has 12 Gb RAM.
On average it has 100-150+ connctions.
At least once a week the firebird server stops responding. Users already connected freezes and no new users can connects.
The easy way to solve this problem is to shutdown the database and restart it. Then it all runs again.
This happen twice yesterday.
When this happens the gab between transactions can be 5000 or it can be 250000.
The clients connects to 7 different terminal servers. From here they connect to the firebrid server.
On thesse 7 terminal servers a lot of other clients connect as well, and the connect to other databases. As it is now we have:
- 1 server running windows and Firebird 1.5 (running 14 Firebird DBs)
- 2 servers running windows and Firebird 2.1.1 (some 10 Firebird DBs)
- 1 server running windows 64 and firebird 2.1.1 64bits (some 9 Firebird DBs)
- 1 linux server (the above) running Firebird 2.1.3 and only 1 DB
All the windows servers with the different firebird versions runs just fine. They never stops responding.
Outside our house we have a lot of setups running. Some uses they own Terminal Server, A firebird server and connect to it as we do.
Others have servers locally and have clients connect locally.
Others just have one machine with all on this.
They use Firebird 1.5 and Firebird 2.1.
None of them has this problem.
The problem is only with the linux server. And this is our biggest DB. Its almost 30 Gb in size now.
In the firebird log we have a lot of these entrys
DBSRV06 (Server) Sun Dec 13 10:02:56 2009
INET/inet_error: select in packet_receive errno = 9
About a year ago we moved this DB from a Windows server because of this exact problem. But we never solved it. It might not happen as often on this linux server, but it happens.
We have come to a point where we do not know where to go from here.
Our client is not happy anymore. They have asked, what we can do.
My colleques has asked if switching to another DB would help (they have mentioned mySQL, Progres, MS SQL and so on). But all of those requieres a lot of reprogramming.
My colleque has been googling some on this problem, and has seen that others experience the same problem.
As of now we have come to the conclusion, that somewhere in Firebird there must be a problem.
But we have no idea what to do now.
If we do not find a solution, is there someone out there that we can pay to help us fix this problem?
I would hate to switch away from Firebird. Besides this customer DB we do not experience any problems, and Firebird runs just fine.
But on the other hand we cant afford to loose this customer.
Regards Michael