present in the cluster. If your client has a priori knowledge that a certain object is stored, it should
retry the request until it succeeds. Normally, only a single retry will be required.
• Network Interface Required. Every DX Storage node requires a working network connection. If a
network cable is unplugged or if the network is not operational at any time during or after startup,
neither SNMP nor SCSP are available. Therefore, the only indication of this condition is in the
attached console (if there is one), where errors such as "Network Unreachable" display.
If the network failure is temporary, DX Storage recovers and continues running after the
connection is restored. SNMP or ping monitoring should be implemented to verify proper network
connectivity among DX Storage nodes.
• Consumer-grade Network Interface Compatibility. The network interface adapters in some
consumer-grade hardware might not initialize correctly, thus preventing DX Storage from starting.
The use of server-grade network interfaces is recommended in DX Storage nodes.
• HTTP Client Library Limitations. Some HTTP client libraries, including Microsoft .NET
HttpWebRequest and httplib in Python, do not handle the Expect: 100-continue header
properly. As explained in the DX Object Storage Application Guide, a client should include this
header when writing streams larger than 64k in length and wait after sending the initial headers
for a response before sending additional content data to DX Storage. Possible responses at that
point are a redirect (301 or 307), an error response, or the 100 Continue response, which means
continue sending data now. In general, it is not permissible, after sending a request including
the Expect: 100-continue header, to continue sending data without waiting, or to continue
sending data if the response is anything other than 100-continue is received. These issues have
been manually repaired in the DX Storage Software Development Kit but integrators writing a
non-SDK client should be aware of these limitations.
• Available Disk Space. The available disk space reported by the DX Storage Management
Console and SNMP is an accurate estimate of the amount of usable space available on a
volume or node. However, the calculation of this value takes into account a number of internal
considerations that might not be immediately visible to an administrator.
For example, DX Storage reserves space on a volume equal to three times the size of the largest
stream stored on a given volume to allow for continuous defragmentation. This means the very
first stream stored on a volume appears to consume three times as much space as necessary
to store the stream itself because, at that time, the very first stream is the largest stream on the
volume.
• Available Index Slots. The management console may over-estimate by a small amount the
number of available index slots in a node. For capacity planning purposes, please use the
estimates found in the memory table of the DX Object Storage Administration Guide.
• Retire in Small Clusters To retire a node or volume, there must be at least two suitable nodes in
the cluster that have storage space available. Volume-less nodes and nodes that are themselves
retired or retiring do not count as suitable nodes. In a multi-server configuration, suitable nodes
can include other nodes running in the same physical chassis as the retiring node or volume.
Copyright © 2010 Caringo, Inc.
All rights reserved
5
Version 5.0
December 2010
ページを印刷する