[ZBXNEXT-1395] Loose agent because of the timeout in agent mounted disk discovery reply Created: 2012 Aug 31  Updated: 2014 Nov 12

Status: Reopened
Project: ZABBIX FEATURE REQUESTS
Component/s: Agent (G)
Affects Version/s: 2.0.0, 2.0.2
Fix Version/s: None

Type: Change Request Priority: Minor
Reporter: Maxim Semenov Assignee: Unassigned
Resolution: Unresolved Votes: 2
Labels: floppy, timeout, windows
Remaining Estimate: Not Specified
Time Spent: Not Specified
Original Estimate: Not Specified
Environment:

windows os with floppy drive



 Description   

Zabbix Agent windows Mounted filesystem discovery problem

If I enable this feature zabbix falls to network errors.
[CODE] 1386:20120829:180101.758 enabling Zabbix agent checks on host [host]: host became available
1382:20120829:180107.819 Zabbix agent item [vfs.fs.discovery] on host [host] failed: first network error, wait for 15 seconds
1386:20120829:180125.933 Zabbix agent item [vfs.fs.discovery] on host [host] failed: another network error, wait for 15 seconds
1386:20120829:180144.032 Zabbix agent item [vfs.fs.discovery] on host [host] failed: another network error, wait for 15 seconds
1386:20120829:180202.144 temporarily disabling Zabbix agent checks on host [host]: host unavailable[/CODE]

It looks like that this is timeout issue.
Is it possible to grow timeout time for low level discovery?

When I run zabbix_get from command line it takes about 6 seconds to get data from host.

[B]WORKAROUND:
Timeout occurs in case of floppy disk presence. I have disabled floppy disk drive in system and timeouts gone away.
Hope this knowledge will help somebody else[/B]

It would be wonderful to skip disk A: or have ability to increase timeout for discovery.



 Comments   
Comment by Alexei Vladishev [ 2012 Aug 31 ]

There is no way to increase timeout for discovery items only. Note that all items (discovery or not) process timeout in exactly the same manner. You may increase Timout parameter in zabbix_agentd.conf and zabbix_server.conf, but it is absolutely not recommended solution.

I am closing it because it is not a bug report.

Comment by Maxim Semenov [ 2012 Aug 31 ]

The problem is:
If you activate "Mounted filesystem discovery"(from default template) for the hosts with floppy disk drive and Windows installed, then we start to get messages like "Zabbix agent on HOST is unreachable for 5 minutes"

This problem because of the function GetVolumeInformation (diskspace.c file) trying to get info about floppy disk drive filesystem. It gets about 5-6 second, so we get timeout.
Think that it's a good idea to exclude disk "a:" from filesystem check in agentd.

Comment by Maxim Semenov [ 2012 Aug 31 ]

The problem is:
If you activate "Mounted filesystem discovery"(from default template) for the hosts with floppy disk drive and Windows installed, then we start to get messages like "Zabbix agent on HOST is unreachable for 5 minutes"

This problem because of the function GetVolumeInformation (diskspace.c file) trying to get info about floppy disk drive filesystem. It gets about 5-6 second, so we get timeout.
Think that it's a good idea to exclude disk "a:" from filesystem check in agentd.

This is not a bug, bun if you want to add file system discovery to a number of windows servers with floppy drives, you will get a lot of "Zabbix agent on HOST is unreachable for 5 minutes" problems. And this is not good.

Comment by richlv [ 2012 Aug 31 ]

i'm not quite sure whether it's a good idea to hardcode such exceptions, especially given the demise of floppy drives. nevertheless, moving to feature requests first

Comment by Oleksii Zagorskyi [ 2012 Nov 08 ]

Just a link - a forum thread (in Russian) http://www.zabbix.com/forum/showthread.php?p=120902
http://www.zabbix.com/forum/showthread.php?p=124056

Comment by Antonio Salazar [ 2013 May 20 ]

I don't think this should be triaged as a minor feature request.

vfs.fs.discovery is not a simple item, but a metaitem where actual items are derived from. The file system discovery always timing out if certain devices are installed invalidates the purpose of autodiscovery because no storage item will be created, this makes file system autodiscovery unreliable.

It's telling that you have to disable storage devices to make storage autodiscovery work.

Comment by Antonio Salazar [ 2013 May 20 ]

I've found file system discovery works if the item type of the discovery rule is changed from "zabbix agent" to "zabbix agent (active)", even if the manual for 2.2 states

The type of check to perform discovery; should be Zabbix agent for file system discovery.


Maybe the documentation should mention this special case?

This makes sense as the server timeout is only for passive checks, active checks can take all the time they need. I guess all hosts that can be passive checked are capable of active checks, so this issue is not that serious after all.

Comment by richlv [ 2014 Apr 12 ]

note that manual has been updated to mention both 'normal' (passive) agent & active agent as valid choices for this discovery rule

Comment by richlv [ 2014 Apr 12 ]

additionally, ZBX-4284 talks about having an algorithm to disable only the items that are timing out. while it won't solve the main problem here, it would prevent host from being reported as being down

Comment by Oleksii Zagorskyi [ 2014 Nov 12 ]

features used in ZBXNEXT-2006 might be considered to improve current case, if it still will be actual.

Generated at Tue Apr 23 12:12:24 EEST 2024 using Jira 9.12.4#9120004-sha1:625303b708afdb767e17cb2838290c41888e9ff0.