Coda File System

CODA or AFS? Read-Only Operation in Production.

From: Ken Price <>
Date: Thu, 17 Jul 2003 11:35:35 -0400
I presently have several web/app servers accessing a common NFS server.
Instead of setting up a replicated/failover NFS server and using a
Heartbeat/DRBD solution, I'd rather save the $$ and replace the NFS server
with a single CODA server (IBM xSeries, Dual PIII, 2Gb RAM, RAID1 SCSI).  My
primary concerns are with regards to disconnected operation, stability, and

1)  Disconnected Operation:  If the CODA server dies (or I reboot it), will
the client machines still keep working off their locally cached copies of
the files?  Will the client have a complete mirror of the files cached
locally, or only the ones already accessed by the webserver?  What if a
rarely hit HTML page is requested for the first time when the CODA server is
unavailable?  When the CODA server comes back online, are the cached files
still available to the web server while CODA checks for updates?

2)  Stability:  I see this asked alot in the archives, and there are always
replies along the lines of, "Yes, but only in this XYZ configuration."  So
I'll ask about this particular scenario:  A Read-Only CODA server and 1-8
high power web and application servers pushing content at the rate of
8-15Mbps and growing.  "Stability" can be defined as an extremely rare
number of sync errors, or ones that can be automatically recovered from.
This is a production environment which should be somewhat self mending.  I
shouldn't have to run integrity checks every day for small syncing errors or
the like.

3)  Robustness:  How efficient is a CODA solution?  I played with it back in
October of last year, and my two 650Mhz Athlon machines with 256Mb RAM were
pushed hard under only moderate load when running as both RW client/servers.
Is a dedicated RO server with dedicated clients going to utilize processor
and memory resources more efficiently?  I'd like the server to function in
other capacities as well (monitoring, low-traffic MySQL server), so even
with 2Gb memory, I see CODA requiring alot of RAM just for itself.

In this scenario, do the members of this list see any major advantages of
using CODA as it stands presently (6.0.1?), over the current version of
OpenAFS (1.2.9a) on a RedHat 9 machine?  Since I am only talking Read-Only
operation, would InterMezzo make any sense?  In a RW environment, it's
terribly buggy and not suitable for production use.

I thank all of you in advance for your replies.


This message is intended only for the use of the addressee and may contain
information that is PRIVILEGED and CONFIDENTIAL and is covered by Mutual Non
Disclosure Agreements, if applicable. If you are not the intended recipient,
dissemination of this communication is prohibited. If you have received this
communication in error, please erase all copies of the message and its
attachments and notify the sender immediately.
Received on 2003-07-17 11:37:50