Optim to make use of REST API / S3 protocol instead CIFS protocol

The problem statement is below:

OPTIM is unable the read any archive file from HCP device (cifs protocol used) and this has become more frequent for the past two weeks. The archive file written locally can be retrieved without any IO Error.

The Optim archiving process currently writes on HCP device (version 9.3 3.14) that makes use of the cifs protocol. The PROD environment being used is OPTIM 11.3.0.8 64-bit on VM Windows Data Center 2016.

OPTIM is severely under pressure in reading the archive data from HCP, returning an IO Error on all our written archive files.

Example as below:

IO error on \\cifs.hcp01-105w.africa.nedcor.net\Optim_OptimCentera\data\VOLANTE_17032022SEQ25096.AF.

OSIO Error in OSIO_ReadFile called from OSIO_ReadSegmentTrailer The operation completed successfully. (00000000)DDName = [None] - Filename = \\cifs.hcp01-105w.africa.nedcor.net\Optim_OptimCentera\data\VOLANTE_17032022SEQ2 5096.AF - position = 412479

This impacts all the Business areas as well as our daily operations whereby we can only archive the data bit not delete them, because of the IO Error on the input archive file.

Also Business are currently experiencing a poor performance while executing the queries they used successfully the past week.

The work around at the moment is, while waiting for the issue to be resolved (replacing cifs protocol with Rest API / S3 protocol), to archive data locally (on our F:\ drive size 400GB).

After a series of discussions with the SAN Storage Team for the use of HCP cifs protocol, they strongly recommend that we make use of Rest API / S3 Protocol since the cifs one is made for small volume of data. Hence, it cannot handle the flow. For your record, we've reported this issue in 2020 and below please find the feedback received from IBM Support L3 (Case TS002561160):

1. IBM do not have the REST API at this time to move data to and from the Hitachi

2. OPTIM is not going to be using the REST API to write archive files to HCP in the foreseeable future

For this to come to pass, The Request for Enhancement (RFE) has to be logged.

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Now, since the root cause is CIFS and although we were using it and it was working before its not an ideal method. Most of the applications hosted on HCP are using REST except for Optim.

We need to seriously look at moving towards REST API for Optim, with Rest we will get more stability, handling a huge volume of data.

We are looking forward to hearing from a quick and effective solution the moment you receive this R.F.E.

Best regards,

Engo.