Enterprise Engineering Portfolio — Abhinav Shrivastava

← Home

Enterprise Clients

18+

Named orgs supported

Industries

Banking, Tech, Gov, Retail…

Geographies

10+

US, India, EU, AU, APAC

Escalations

25+

Production issues resolved

Severity

P1/P2

Most were production-down

This portfolio documents named enterprise customers whose production issues I directly investigated, root-caused, and resolved as a storage and compute engineer on IBM PowerVC — an OpenStack-based private cloud platform for IBM Power Systems.

A "customer escalation" in this context means the issue was reported from a live production environment, filed through IBM's enterprise support system, and assigned to the engineering team for a code-level fix or targeted patch. Many were P1 (critical production outage) or P2 (major function impaired) severity — requiring fast turnaround and direct customer communication. Customer identities have been generalised to protect confidentiality while preserving the industry, geography, and scale context.

Banking & Financial Services

PSB

Large Private Sector Bank (India)

Private sector bank · India · Multiple engagements

NetworkFibre Channel port issue preventing VM deployment — FC ports misconfigured, storage database lookup failing
NetworkZones deleted incorrectly when using initiator-vfc zoning policy — skip processing if WWPN absent from DB
NetworkZone name template truncation for dual initiators in initiator-vfc policy causing connectivity failures
StorageFC port configuration page — Fibre Channel ports not discovered in port configuration view

PUB

Public Sector Bank (India)

Public sector bank · India · IPG migration issue

ComputeVM live migration failure — VMs associated with Initiator Port Groups (IPGs) did not list any eligible destination hosts in migration dialog

NAT

Nationalised Bank (India)

Public sector bank · India · Zoning enforcement

NetworkStorage template-defined FC ports not being used for zoning with LPAR adapters — platform auto-selecting wrong ports instead of honoring template
StorageUnable to map LUN to multiple hosts from a specific server — multipath LUN mapping failure

MCB

Major Central Bank (US)

Central bank · United States · Image import failure

ImagePost-upgrade image import failure — Python logic bug prevented importing VM images with more than one data volume after upgrade to PowerVC 2.3.2

CBK

Commercial Bank (South Africa)

Commercial bank · South Africa · Storage deletion

StorageCustomer unable to delete a volume on Hitachi storage — volume stuck in error state, blocked by residual host mappings on the storage array

Technology & Software

GTC

Global Technology Company (US)

Technology · United States · Multiple P1/P2 escalations

StorageCinder API worker lock contention — Hitachi storage driver holding global lock between API and volume workers causing log pipeline stall (Part 1 & 2)
StorageData volumes stuck in "Attaching" state — attach flow not completing, requiring manual state recovery
PerfLKU (Live Kernel Update of host OS) with 100 volumes — bulk attach timed out on surrogate partition; attach took longer than SLA
StorageIntermittent 503 Service Unavailable errors in storage API calls
StorageBulk detach failed with auth connectivity error (keystoneauth1 ConnectFailure)
StorageHost mapping visible on storage array while volume shows Available in UI — volume deletion should be blocked until host mapping is cleared

GRE

Global Retail Enterprise (Sweden)

Retail / Technology · Sweden · HPE storage issues

StorageHPE Alletra storage registration failure — platform unable to register the storage array
StorageHPE storage driver missing multipath flag in validate_connection_type() causing single-path connectivity
StorageHPE storage in ERROR state with stack traces in cinder-volume logs after registration
ImageImage export/import fails on HPE storage — incorrect handling of absent initiator in terminate_connection

EOV

European Observability Software Vendor

Observability software · Austria · Upgrade regression

ImageVM images stopped working after upgrade from PowerVC 2.2.1.2 to 2.3.1 — regression in image metadata processing caused all image-based deployments to fail

TSF

Technology Services Firm (US)

Technology services · United States · Zoning anomaly

NetworkZonemanager randomly doubling zone count and paths for a VM when adding a new LUN — race condition in FC zone management caused duplicate zones across storage fabric

Retail & Consumer

MFR

Major Fashion Retailer (US)

Fashion retail · United States · Major feature driver

StorageDrove the entire Initiator Port Group (IPG) feature implementation — DB schema, API, upgrade scripts, CLI display tool, boot volume tagging, HMC & NovaLink deploy flows, onboard/manage edge cases
StorageIPG tag for boot volumes during Storage Connectivity Group creation — ensures correct storage path affinity for IBMi workloads
PerfBulk attach SVT script fixes and formatting for high-volume attach test scenarios

GBC

Global Beverage Conglomerate (Brazil)

Beverage / FMCG · Brazil · Volume attachment failures

StorageVolume detach failure — KeyError in data response from storage connector left volume stuck in detaching state
DBVolume attachment interruption left 14 orphaned BDM (block device mapping) entries in database — customer could delete volumes but stale DB records remained, blocking future operations (PowerVC 2.3.2)

Government & Public Sector

GIA

Government IT Agency (Australia)

Government IT services · Australia · Multiple deep-dive engagements

ComputeVMs deployed with only 50% of expected storage paths — IPG / SPG (Storage Port Group) path allocation halved during VM deployment
StorageRemoving a compute host should delete associated Initiator Port Groups — platform leaving empty IPGs behind after host removal
Storagemovevdisk retry logic — SVC storage operation intermittently failing, needed retry wrapper
StorageReplaced FC port handling in powervc-capture/restore scripts — IPG restore scripts not accounting for WWPN port replacement events
StorageAuthored powervc-capture and restore scripts for IPG (Initiator Port Group) backup and recovery

Managed Services & IT Providers

MSP

Managed IT Services Provider (US/Europe)

IT services · US & Europe · Multi-release fixes

NetworkVM deletion intermittently failing to remove Fibre Channel fabric zones and host objects on storage array
ComputeRemote Restart (RR) / Automated Remote Restart (ARR) failing for VMs with IPG enabled — HA restart not honoring port group assignments on 2.1.1 + ifix

ISP

IT Solutions Provider (Canada)

IT services · Canada · Stale volume attachment

DBDeleted volume still showing as attached to a VM — orphaned attachment record in database after volume deletion, causing inconsistent state visible to customers

EIT

Enterprise IT Services Company (Finland)

IT services · Finland · Upgrade documentation gap

ImageDocumentation gap: customers not warned to update ALL backup nodes before upgrading to PowerVC 2.3.1 — RPC version mismatch between nodes prevented image imports after partial upgrade

Enterprise & Other

EUC

Enterprise Upgrade Customer

Enterprise · Upgrade issue

StoragePost-upgrade to PowerVC 2.3.1 — incorrect error message displayed when a storage template is not set on a volume, making diagnosis harder for operators

FSF

Financial Services Firm (US)

Financial services · United States · Bulk ops & replication

StorageBulk volume detach never completing on IBM PowerVS environment — detach pipeline hung indefinitely
StorageAuxiliary volume size change not reflecting after 24 hours — global mirror replication volume sizing out of sync after resize operation
StorageGRS (Global Replication Service): Auxiliary volume changes for failover/failback flows

SVP

Storage Vendor (POC Engagement)

Storage vendor · VM capture integration

ImageVM capture failing with Glance (image service) HTTP 500 error during PowerVC 2.3.0 integration validation for new storage backend

IED

Internal Enterprise Deployments

IBM internal enterprise deployments · Multiple P1 issues

ComputeLKU (Live Kernel Update) failures across multiple customer deployments — surrogate LPAR creation failing with mixed NPIV/SSP environments
StorageUnable to manage hosts with mixed NPIV and SSP storage connectivity
StorageUnable to add IBM Storwize V7000 as pluggable storage provider
ComputeCluster state changing to Error / Critical after HMC firmware update

CPV

Consumer Products Company (US/PowerVS)

Consumer products · US · PowerVS SRE escalation

PerfSlow response to volume detach in IBM PowerVS environment — SRE-reported latency issue in detach pipeline under high concurrency

GEN

Generic Enterprise (unnamed field reports)

Multiple enterprise clients · Field-reported production issues

StorageVolume page slow to load with large number of volumes — UI refresh performance degradation
StorageHitachi storage spinning forever due to bad/expired TLS certificate — no timeout handling
ComputeHMC idle session leak — PowerVC leaving open sessions on Hardware Management Console over time
StorageStorage connectivity group broken after network change
StorageStorage in Error/Critical state — unexplained transition to error after routine operations
DBpowervc-dbcheck utility fails to recognize non-zero deleted field values in various database tables

Issue Category Summary

Category	What it means	Issues
Storage	Volume create/attach/detach/delete failures, SAN array driver bugs, host mapping issues	18
Network	Fibre Channel zone management, port discovery, path count issues	8
Compute	VM lifecycle (deploy, migrate, restart), hypervisor connectivity, host management	7
Image	VM image export/import/capture failures, upgrade regressions on image workflows	6
DB	Orphaned database records, stale attachment state, consistency checker gaps	4
Perf	Bulk operation timeouts, API response latency, attach/detach throughput regressions	4