Event ID 51 Source=Disk

[jlt22]

Getting the error mentioned above in my system log on w2k3R2 server. The Description on the error is "An error was detected on device \Device\Harddisk2 during a paging operation."

This error is repeated a few times then I get an "application popup" event id 26 with the following description reported "Application popup: Windows - Delayed Write Failed : Windows was unable to save all the data for the file E:\Logs\Kiwi. The data has been lost. This error may be caused by a failure of your computer hardware or network connection. Please try to save this file elsewhere.

I also get an event id 50 mixed in there Source=NTFS with the following description {Delayed Write Failed} Windows was unable to save all the data for the file . The data has been lost. This error may be caused by a failure of your computer hardware or network connection. Please try to save this file elsewhere.

This bundle of errors has happend twice in the past 24 hrs and lasts for approx 7 minutes, then things calm down. The last error in both sequences where this happened was the event id 26 but the path to the data where it was unable to save was E:\$MFT.

Any ideas on how to nail down the source of this problem?

[aaron (staff)]

This means the connection was broken or was unresponsive for quite long period of time. Double-ckeck your network hardware, try to stress it with enormous load. Let us know about the results. If everything will be fine we'll provide you with log-enabled version of the StarPort as it's the last one to blame in the whole chain. Thanks!

Getting the error mentioned above in my system log on w2k3R2 server. The Description on the error is "An error was detected on device \Device\Harddisk2 during a paging operation."

This error is repeated a few times then I get an "application popup" event id 26 with the following description reported "Application popup: Windows - Delayed Write Failed : Windows was unable to save all the data for the file E:\Logs\Kiwi. The data has been lost. This error may be caused by a failure of your computer hardware or network connection. Please try to save this file elsewhere.

I also get an event id 50 mixed in there Source=NTFS with the following description {Delayed Write Failed} Windows was unable to save all the data for the file . The data has been lost. This error may be caused by a failure of your computer hardware or network connection. Please try to save this file elsewhere.

This bundle of errors has happend twice in the past 24 hrs and lasts for approx 7 minutes, then things calm down. The last error in both sequences where this happened was the event id 26 but the path to the data where it was unable to save was E:\$MFT.

Any ideas on how to nail down the source of this problem?

[jlt22]

I think I am doing that already, the same switch gets hammered every night when backups run, have not had this error during backup phase. Also, as for stressing the disk side, I have a kiwi syslog service running on this server and am getting consistant feeds of snmp traffic from 24 firewalls writing logs to this disk 24X7.

Coraid SR1521 connected via 1GB port to Dell PE 1750 running W2K3R2. Currently there is a dell GB switch in between with vlan to segment from other network traffic. We just happen to be replacing the Dell switch this week with a Foundry switch which will eliminate one piece of hardware, although I don't think the switch is the problem.

As for logging, the Starport.log file located under user profile\local settings\application data does not seem to be using the same time zone as the clock on the server, is there a way to change this? Also, it would be helpful to have the date recorded in this log file if you have not incorporated it into your latest release. I am awating replacement license files before I can update the starport drivers so disregard the last comment if you have changed the format of the log file to include the date.

Please let me know if there is a way to turn up logging with current starport version to try to find the cause of the problem or to get better info than the windows disk error.

[aaron (staff)]

1) You need to stress the hardware not with the StarPort. That's the point...

2) Need latest driver binary in any case.

I think I am doing that already, the same switch gets hammered every night when backups run, have not had this error during backup phase. Also, as for stressing the disk side, I have a kiwi syslog service running on this server and am getting consistant feeds of snmp traffic from 24 firewalls writing logs to this disk 24X7.

Coraid SR1521 connected via 1GB port to Dell PE 1750 running W2K3R2. Currently there is a dell GB switch in between with vlan to segment from other network traffic. We just happen to be replacing the Dell switch this week with a Foundry switch which will eliminate one piece of hardware, although I don't think the switch is the problem.

As for logging, the Starport.log file located under user profile\local settings\application data does not seem to be using the same time zone as the clock on the server, is there a way to change this? Also, it would be helpful to have the date recorded in this log file if you have not incorporated it into your latest release. I am awating replacement license files before I can update the starport drivers so disregard the last comment if you have changed the format of the log file to include the date.

Please let me know if there is a way to turn up logging with current starport version to try to find the cause of the problem or to get better info than the windows disk error.

[jlt22]

No need to be condescending. I did mention that the switch between the server and the coraid box gets hammered every night during backup. And that while the backup is happening I have not seen the problem on the server in question. So while the backups are going on and hammering the same switch, the server in question is writing its log files via starport to the coraid box and doesn't miss a beat when the switch is really busy with other traffic.

I would call this stressing the hardware "not with starport", at least the switch.

How would you recommend I stress the coraid box and not use starport?

[aaron (staff)]

1) Check your network. No StarPort, no CoRAID box. Only NIC on the server and switch put under heavy load.

2) Make sure it's not CoRAID inserting delays (it should keep it's own logs).

No need to be condescending. I did mention that the switch between the server and the coraid box gets hammered every night during backup. And that while the backup is happening I have not seen the problem on the server in question. So while the backups are going on and hammering the same switch, the server in question is writing its log files via starport to the coraid box and doesn't miss a beat when the switch is really busy with other traffic.

I would call this stressing the hardware "not with starport", at least the switch.

How would you recommend I stress the coraid box and not use starport?